Skip to main content

Full text of "The photosynthesis of carbon compounds"

See other formats

The photosynthesis 
of carbon compounds 


Photosynthesis of 
Carbon Compounds 

Melvin Calvin 
J. A. Bassham 

University of California 
Berkeley, California 

W. A. Benjamin, Inc. 

New York 



Copyright © 1962 by W. A. Benjamin, Inc. 
All rights reserved 

Library of Congress Catalog Card Number: 62-10567 
Manufactured in the United States of America 

The manuscript was received November 15, 1961, and published 
February 27, 1962. 


2465 Broadway 

New York 25, New York 




Nearly sixty years ago, Emil Fischer described his experi- 
ments which led to the discovery of the structure of glucose 
and related sugars. In the past fifteen years Melvin Calvin 
and his associates have performed experiments leading to an 
understanding of the reactions used by photosynthetic organ- 
isms to make these sugars and many other compounds from 
carbon dioxide, water, and minerals, using the energy of light. 

It was not long after the basic reaction of photosynthesis 
was recognized that speculation regarding its mechanism com- 
menced. These discussions were carried forward first by Justus 
von Liebig and then by Adolf von Baeyer and, finally, by 
Richard Wilstatter and Arthur Stoll, into this century. How- 
ever, it was the mechanism of the reverse pathway, that is, 
the combustion of carbohydrate to carbon dioxide and water 
with the utilization of the energy, which was first successfully 
mapped. This pathway was elucidated primarily by Otto 
Meyerhof and Hans Krebs. 

Professor Calvin's interest in the basic process of solar 
energy conversion by green plants began about 1935, when he 
was studying with Professor Michael Polyani at Manchester, 
There he became interested in the remarkable properties of 

coordinated metal compounds, particularly metalloporphy- 
rins, as represented by heme and chlorophyll. He began a 
study on the electronic behavior of such compounds at that 
time. When Professor Calvin joined the Chemistry Depart- 
ment at Berkeley, these studies were encouraged by Professor 
Gilbert N, Lewis, and they have been continued to the present 
time. In time they will contribute to our understanding of the 
precise way in which chlorophyll and its relatives accomplish 
the primary quantum conversion into chemical potential, 
which is then used to drive the carbohydrate synthesis. 

It has long been known that the reduction of carbon di- 
oxide to carbohydrate is probably a dark reaction, separate 
from the primary quantum conversion act. This knowledge 
stemmed from the early work of F. F. Blackman on the dark 
reactions of photosynthesis and its interpretation by Otto 
Warburg, and particularly from the comparative biochemical 
studies of Cornelius van Niel. Finally, Robert Hill separated 
the photo-induced production of molecular oxygen chemi- 
cally and physically from the reduction of carbon dioxide 
when he demonstrated oxygen evolution by illuminated 
chloroplasts, using ferric ion as an oxidant in place of carbon 

We can summarize the over-all conversion of light energy 
into chemical energy in the form of carbohydrate and 
oxygen by several steps. First, the light energy absorbed by 
chlorophyll and related pigments is converted into the high 
chemical potential energy of some compounds. Second, these 
compounds react with water and produce oxygen and good 
reducing agents as well as other cofactors containing high 
chemical f)otential energy. Finally, these reducing and ener- 
getic cofactors react with carbon dioxide and other inorganic 
compounds to produce organic compounds. 

One of the principal difficulties in studying the synthetic 
pathway is that the machinery which converts carbon dioxide 
and minerals to organic compounds is itself composed of 
organic compounds made up of the same elements. Ordinary 
analytical methods do not allow us to distinguish easily be- 


tween the machinery and its substrate. Fortunately, the dis- 
covery of the long-lived isotopic carbons (carbon- 14) by 
Samual Rubin and Martin Kamen in 1940 provided the 
ideal tool for tracing these synthetic routes. 

In 1945, carbon- 14, radiocarbon, became available in 
large amounts as a product from nuclear reactors. With the 
encouragement and support of Professor Ernest O. Lawrence, 
the Director of the Radiation Laboratory in Berkeley, Pro- 
fessor Calvin began to study the pathway of carbon reduction 
during photosynthesis, using carbon- 14 as his principal tool. 

Among a number of people who were to be associated 
with him during the next few years of this work and who 
would all contribute to the success of the research. Dr. Andrew 
A. Benson was particularly instrumental, especially in the 
identification of the early products of photosynthetic carbon 
reduction. Key contributions to the development of the 
carbon reduction cycle were made by Dr. Peter Massini and 
Dr. Alex Wilson. Beginning as a graduate student with Pro- 
fessor Calvin in 1947, I have had the pleasure of being asso- 
ciated with him in this work to the present time. 

The first big success came with Professor Calvin's identi- 
fication of phosphoglyceric acid as the first stable product of 
carbon reduction during photosynthesis. Soon thereafter the 
application of two-dimensional paper chromatography com- 
bined with radioautography became an invaluable analytical 
tool for separating the minute amounts of radioactive ingre- 
dients formed in the plant. Identification of the remaining 
intermediates in the carbon reduction cycle soon followed, 
and these turned out to be all sugar phosphates. 

A combination of kinetic studies on the appearance of 
carbon- 14 in these intermediates, with degradation of the 
compounds that revealed the location of the radiocarbon in 
individual atoms, soon led to a linking together of a reaction 
sequence leading from phosphoglyceric acid through the 
several sugar phosphates. The experiments of Massini and 
Wilson helped to establish the carboxylation and reduction 
reactions of photosynthesis, and the cycle was complete. 


In succeeding years much work has been done to check 
the validity of the cycle, to investigate details of its mecha- 
nism, and to establish its quantitative importance. 

From almost the beginning of these studies we have been 
interested in reactions leading from the cycle to various other 
synthetic intermediates and end products, such as amino acids, 
sucrose and polysaccharides, and carboxylic acids. As a result 
of this work we have found that the photosynthetic machine, 
the chloroplast, is an even more complex and diversified ap- 
paratus than had been suspected. Not only does it manufac- 
ture sugars and other carbohydrates, but apparently nearly all 
other organic materials necessary for its continued growth as 

In this book we review the evidence leading to the formu- 
lation of the carbon reduction cycle and discuss its quantita- 
tive importance. We describe as far as possible the biosynthetic 
pathways which we believe exist in the chloroplast. We show 
how newly reduced carbon from the carbon reduction cycle 
provides the starting material for these pathways. Our ob- 
jective is to map complete synthetic sequences from carbon 
dioxide to final products. Three papers, of fundamental im- 
portance in the development of the theory regarding the path 
of carbon in photosynthesis, are included as reprints. 

We are now just at the threshold of discovery of many of 
the biosynthetic pathways. There is good experimental evi- 
dence for some and a few clues for others, but for many we 
must speculate, relying on known, but nonphotosynthetic, 
pathways. We have called on our experience of some fifteen 
years' study of carbon fixation patterns during photosynthesis 
to provide us with clues. The clues help us to predict which 
reactions, which pathways, and which intermediates may be 
considered to be likely participants in the photosynthesis of 
carbon compounds. 

This year Professor Calvin was awarded the Nobel Prize 
for his work on the assimilation of carbon dioxide during 
photosynthesis. Those who have worked with him and have 
experienced the stimulation provided by his enthusiasm and 


insight are especially delighted by this most well-deserved 
recognition of one of his many scientific achievements. Those 
of us who, under his leadership, have contributed something 
to the development of the carbon reduction cycle are particu- 
larly pleased to have been a part^iDf this exciting work. 

James A. Bassham 

Berkeley, California 
December 1961 



The publisher and the authors wish to acknowledge the 
assistance of the following organizations in the preparation 
of this volume: 

The United States Atomic Energy Commission, which 
sponsored the preparation of this volume. 

Verlag-Birkhauser A.-G., Basel, for permission to reprint 
the article from Experientia. 

The American Chemical Society, Washington, D.C., for 
permission to reprint the article from the Journal of the 
American Chemical Society. 

Elsevier Publishing Co., Inc., Amsterdam, for permission 
to reprint the article from Biochimica et Biophysica Acta. 



^^^ Contents 

Preface v 

Acknowledgments x 

Introduction 3 

Carbon reduction cycle of photosynthesis 8 

Evidence for the carbon reduction cycle 12 

The carboxylation reactions 21 

Balance among synthetic pathways 25 

Photosynthesis vs. other forms of biosynthesis 27 

Amino acid synthesis 29 

Carboxylic acids 37 

Carbohydrates 49 

Fats * 56 

Pigments 60 

Aromatic nuclei 65 

Other biosynthetic products 67 

References 69 





The Path of Carbon in Photosynthesis: XX. The 
Steady State, by M. Calvin and P. Massini, Experi- 
entia, VIII/12, 445-457 (1952) 79 

The Path of Carbon in Photosynthesis: XXI. The 
Cyclic Regeneration of Carbon Dioxide Acceptor, 
by J. A. Bassham, A. A. Benson, Lorel D. Kay, 
Anne Z. Harris, A. T. Wilson, and M. Calvin, 
;. Am. Chem. Soc, 76, 1760-1770 (1954) 92 

Dynamics of the Photosynthesis of Carbon Com- 
pounds: I. Carboxylation Reactions, by J. A. Bass- 
ham and Martha Kirk, Biochim. et Biophys. Acta, 
43,447-464 (1960) 103 

Index 121 


The photosynthesis 
of carbon compounds 

'^^ Introduction 

Biosynthesis begins with photosynthesis. Green plants 
and other photosynthetic organisms use the energy of ab- 
sorbed visible light to make organic compounds from in- 
organic compounds. These organic compounds are the 
starting point for all other biosynthetic pathways. 

The products of photosynthesis provide not only the 
substrate material but also chemical energy for all subsequent 
biosynthesis. For example, nonphotosynthetic organisms 
making fats from sugars would first break down the sugars 
to smaller organic molecules. Some of the smaller molecules 
might be oxidized with O2 to CO2 and water. These reac- 
tions are accompanied by a release of chemical energy, be- 
cause O2 and sugar have a high chemical potential energy 
toward conversion to CO2 and H2O. In a biochemical system 
only part of this energy would be released as heat. The rest 
would be used to bring about the conversion of certain 
enzymic cofactors to their more energetic forms. These co- 
factors would then enter into specific enzymic reactions in 
such a way as to supply energy to drive reactions in the 
direction of fat synthesis. Fats would be formed from the 
small organic molecules resulting from the breakdown of 

sugars. Thus sugar, a photosynthetic product, can supply 
both the energy and the material for the biosynthesis of fats. 

Photosynthetic organisms achieve energy storage through 
their ability to convert electromagnetic energy to chemical 
potential energy. The conversion begins when pigments 
absorb light energy. The absorbed energy changes the elec- 
tronic configuration of the pigment molecule (chlorophyll) 
from its ground energy state to an excited state. The return 
of the pigment molecule to its ground-state energy level is 
accompanied by a (chemical) reaction that would not proceed 
without energy input; i.e., the products of this reaction have 
a smaller negative free energy of formation from their ele- 
ments than do the reactants (in the same reaction). Thus some 
of the light energy is converted to chemical potential. 

The detailed mechanism of all these energy-conversion 
steps is not known. However, the net result is often formu- 
lated by two chemical equations. One of these is an oxida- 
tion-reduction reaction resulting in the transfer of hydrogen 
from water to triphosphopyridine nucleotide (TPN): 

(1) HOH + TPN+ -^ iOa + TPNH + H+ 

AF' = +52.6 kcal* 

The other reaction is the formation of an anhydride, adeno- 
sine triphosphate (ATP), from the ions of two phosphoric 
acids, adenosine diphosphate and orthophosphate: 

(2) ADP3- + HP04= ^^ HOH + ATP^" + H + 

AF' = +11 kcal* 

In each of these reactions some of the light energy is stored 
as chemical potential, as indicated by the positive quantities 
for free energy change. 

The structural formulas of these two cofactors are shown 
in Figure 1. TPNH and its close relative DPNH (reduced 
diphosphopyridine nucleotide) serve a double function in 
photosynthesis and in all biosynthesis. Both TPNH and 

* Assuming these concentrations: (TPNH) = (TPN + ), 
(ATP*-) = (ADP-^-), (H + ) = 10-' M, (HP04=) = 10-=^ M. 
















HC — — P- 













N C — N, 


N I 












Triphosphopyridine nucleotide (oxidized form) 


HC — O — P-0-P-O-P-OH 

Adenosine triphoiphote (ATP) 

In Adenosine diphosphote (AOP), 
terminal phosphate is replaced by -OH. 





NIcotinomide portion of 
TPNH (reduced TPN+) 

Figure 1. Formulas of TPN and ATP. 

DPNH are reducing agents and carriers of chemical poten- 
tial, in other words, strong reducing agents. Thus, one of 
their roles in biochemistry is analogous to that of H2 in syn- 
thetic organic chemistry. 

The function of ATP is to carry chemical potential and 
to act as a powerful phosphorylating agent. In the reduction 
of an acid to an aldehyde, important in photosynthesis, its 
role may be compared to that of a mineral acid anhydride 
in organic synthesis: 

Organic synthesis: 

Car- Acid 

boxylic anhy- 

acid dride 


(3) R— C + iPCI, 








— > 4H3P03 





R— C -f HCl 



Carboxylic Acid Reducing 

acid anhydride Acyl derivative agent 

o o 

/ /■ TPNH 

(4) R— C + ADP— O— PO3H I > R— C > 

\ \ enzyme 

O- OPO3H- 

> ADP 

Aldehyde Acid 


R— C + HOPO3H- 



Among the many other reactions of ATP in biosynthesis, one, 
which is of considerable importance in photosynthesis, is the 
formation of sugar phosphates from sugars: 

(5) H+ + ROH + ADP— O— POgH" -> 

R— OPO3H- + ADP + H2O 

The only known reactions of the carbon reduction cycle in 
photosynthesis which would require the use of TPNH and 
ATP are of the type shown in Eqs. (4) and (5). These re- 
actions are the means by which chemical potential, derived 
from the absorbed light, is used to bring about the reduction 
and transformation of carbon from CO2 to organic com- 

These two cofactors, ATP and TPNH, are at present 
the only ones that are known to be generated by the light 
reactions of photosynthesis and at the same time seem to be 
required for steps in the carbon reduction cycle. The possi- 
bility remains, however, that there are other energetic or 
reduced cofactors acting as carriers of hydrogen and energy 
from the light reactions to the carbon reduction cycle. Such 
unknown cofactors might substitute for or replace TPNH 
or ATP. They could, in fact, be more effective than the 
known cofactors, particularly in vivo, where they might well 
be built into the highly organized structure of the chloro- 
plast. If such unknown cofactors do exist, they would have 

to perform essentially the same functions as TPNH and ATP 
and would presumably be about as effective as carriers of 
chemical potential. In all discussions of the role of TPNH 
and ATP, the possibility of their replacement by as-yet- 
unidentified cofactors should be kept in mind. 

For the purpose of discussion, let us consider the photo- 
synthesis of carbon compounds as an isolated set of reactions. 
The principal substrates for this set of reactions are CO2, 
hydrogen (as TPNH), phosphate (as ATP), and NH4 + . The 
ammonium ion may be contained in the plant nutrient or 
it may be derived from the reduction of nitrate. If nitrate 
reduction is the source of NH4 + , the energy for the reduc- 
tion must also come from the light, at least indirectly. Other 
probable inorganic substrates for photosynthesis of organic 
compounds include sulfate, magnesium ion, and a number 
of trace elements. Many of these are required for growth in 
plants but may or may not be incorporated in organic com- 
pounds by photosynthesis. 

^^ reduction cycle 
"^i^r of photosynthesis 

We believe the principal pathways for the photosyn- 
thesis of simple organic compounds from CO2 to be those 
shown in Figure 2 (1,2). The points at which ATP and 
TPNH act in these pathways are indicated. Kinetic studies 
(3) show that these pathways account for nearly all the car- 
bon dioxide reduced during photosynthesis, at least in the 
unicellular algae Chlorella pyrenoidosa. From other inves- 
tigations (4) it appears that the general metabolic sequence 
is the same in most respects for all photosynthetic organisms. 
(We shall discuss the recently proposed role of glycolic acid 
in CO2 reduction in the section on Carboxylic Acids.) 

The central feature of carbon-compound metabolism 
in photosynthesis is the carbon reduction cycle. Most of the 
carbon dioxide used is incorporated via this cycle. Pathways 
lead from intermediates in the cycle to various other impor- 
tant metabolites. A few of these pathways are shown in Fig- 
ure 2. 

The initial step for carbon dioxide incorporation in 
the cycle is the carboxylation of ribulose-l,5-diphosphate at 
the number 2 carbon atom of the sugar to give a highly 

labile ^-keto acid. Evidence for the existence of this unstable 
intermediate has been adduced from in vivo studies (5) . It 
has not been isolated in the in vitro reaction with the enzyme 
carboxydismutase. The product of the reaction in vitro is 2 
molecules of 3-phosphoglyceric acid (PGA). The products 
in intact photosynthesizing cells may be 2 molecules of PGA 
or, as kinetic studies indicate (3), 1 molecule of PGA and 1 
molecule of triose phosphate. 

Once formed, the PGA is transformed in two ways. 
Some molecules are converted to products outside the cycle 
while the remainder are reduced to 3-phosphoglyceraldehyde 
via a reaction of the type shown in Eq. (4). The enzymes 
responsiule for the two successive steps in the reduction are 
probably similar to phosphoglycerylkinase (6) and triose 
phosphate dehydrogenase (7-10). 

The next phase of the carbon reduction cycle is the 
conversion of 5 molecules of triose phosphate to 3 molecules 
of pentose phosphate by a series of reactions. These reactions 
include condensations (aldolase), carbon-chain-length dismu- 
tations (transketolase), removal of phosphate groups (phos- 
phatase), and interconversions of different pentose phos- 
phates (isomerase, epimerase). Enzyme systems that catalyze 
reactions similar to these steps are listed in Table 2. The 
sequence of steps may be seen in the cycle diagram (Figure 2). 

The various pentose phosphates are converted to ribu- 
lose-5-phosphate. The final step is the formation of ribulose 
diphosphate (RuDP) from ribulose-5-phosphate. This step 
requires 1 molecule of ATP [Eq. (5)]. 

For every reaction in the cycle to occur at least once 
(a complete turn of the cycle), the carboxylation reaction 
must occur three times. The net result of each complete 
turn of the cycle is the incorporation of 3 molecules of CO2 
and the production of 1 three-carbon (or V2 six-carbon) or- 
ganic molecule. Each complete turn of the cycle would re- 
quire 6 molecules of TPNH or equivalent reducing cof actor 
(2 per CO2) and 9 molecules of ATP, if each Ce carboxylation 
product is split to 2 molecules of PGA and if all the PGA 




is reduced to triose phosphate. If the carboxylation product 
is reductively split (dashed line in Figure 2) the requirement 
for TPNH would probably be the same, that is, 6 molecules 
per complete turn of the cycle. In this case, however, the 
cycle might require either 9 molecules of ATP or only 6. 

Figure 2. Carbon reduction pathways in photosynthesis. Com- 
pounds: (1) 2-carboxy-3-keto-l,5-diphosphoribitol, (2) 3-phospho- 
glyceric acid (3-PGA), (3) glyceraldehyde-3-phosphate, (4) dihy- 
droxyacetone phosphate, (5) fructose- 1,6-diphosphate, (6) ery- 
throse-4-phosphate, (7) sedoheptulose-l,7-diphosphate, (8) xylu- 
lose-5-phosphate, (9) ribose-5-phosphate, (10) ribuIose-5-phosphate, 
(11) ribulose-l,5-diphosphate, (12) 2-phosphoglyceric acid (2- 
PGA), (13) phosphoenolpyruvic acid (PEPA), (14) oxalacetic acid. 
_@: fructose diphosphate and sedoheptulose diphosphate lose 
one phosphate group before transketolase reaction occurs. 


^<f for the carbon 
'T^ reduction cycle 

The carbon reduction cycle in essentially the form 
shown in Figure 2 was mapped during the period between 
1946 and 1953 (11-17). The experiments, results, and inter- 
pretations leading to its formulation have been extensively 
discussed elsewhere (2). They will be briefly reviewed here, 
not necessarily in chronological order. 

The carbon that enters the plants' metabolism has been 
followed through the various intermediate compounds by 
labeling the carbon dioxide with radiocarbon, C^*. The 
analysis of the labeled compounds has been carried out by 
paper chromatography and radioautography. The interpre- 
tation of results leading to the cycle formulation has been 
based on the kinetics of the appearance of C^* in various 
identified compounds as a function of time of photosynthesis 
with C^*02 and other variables. 

The methods are best described by an illustration. Con- 
sider a simple experiment with a suspension of the algae 
Chlorella pyrenoidosa, very extensively used in these studies. 
These green unicellular plants, suspended in water contain- 
ing the necessary inorganic ions (nitrate, phosphate, etc.) 
and aerated with a stream of C^-Oo (ordinary carbon dioxide), 


photosynthesize at a rapid rate if illuminated from each side 
in a thin transparent vessel. The CO2 is continually taken 
up from the solution (where it is in equilibrium with bicar- 
bonate ion) and converted by the photosynthetic plant 
through a series of biochemical intermediates to various 
organic products. 

A solution of radioactive bicarbonate, HC^'^Oa-, is sud- 
denly introduced into the algae suspension. The plant does 
not distinguish in any important way between the C^^ ^nd 
C^*, which are chemically almost identical. Immediately some 
of the C^* is incorporated into the first of the biochemical 
intermediate compounds. As time passes the C^* gets into 
subsequent intermediates in the chain. After a few seconds 
exposure to the C^^02, the suspension of algae is run into 
methanol to a final concentration of 80 per cent methanol. 
This treatment denatures all the enzyme instantly and freezes 
the pattern of C" labeling by preventing further change. 
Now the dead plant material is analyzed for radioactive com- 
pounds to see which are the first stable products of carbon 
reduction during photosynthesis. 

The first step in this analysis is to prepare an extract 
of the soluble compounds. The early products of carbon 
reduction have been found to be simple soluble molecules. 
This extract is then concentrated and analyzed by the method 
of two-dimensional paper chromatography (12). The impor- 
tance of the method for these studies stems from the fact that 
it permits the analysis of a few micrograms or less of dozens 
of different substances in a single simple operation. 

Of these many compounds, those into which the plant 
incorporates C^^ during its few seconds of photosynthesis 
with HC^'^Os- are radioactive and omit the particles result- 
ing from radioactive decay of the C^". In this case these are 
13 particles, and these may be detected by the fact that they 
expose x-ray film. Thus, if a sheet of x-ray film is placed in 
contact with the two-dimensional paper chromatogram, sub- 
sequent development of the film will show a black spot on 
the film corresponding to the exact shape and location of 


each radioactive compound on the paper. A quantitative 
determination of the amount of radiocarbon in each com- 
pound may then be made by placing a Geiger-Miiller tube 
with a very thin window over the radioactive compound on 
the paper and counting electronically the emitted ^ particles. 

The next stage in the method of radiochromatographic 
analysis is the identification of the radioactive compounds. 
This identification is accomplished in a variety of ways. 
When a familiar set of chromatographic solvents has been 
used, the position of an unknown compound compared to 
the positions of known substances provides a clue to its iden- 
tity. The next step may be elution or washing of the com- 
pound off the paper and the determination of such chemical 
and physical properties (e.g., the distribution coefficient) of 
the substance as can be measured with a solution of a few 
micrograms or less of the material. These properties are then 
compared with those of known compounds. The final check 
on the identity of the compound is frequently made by plac- 
ing on the same spot on filter paper the radioactive com- 
pound and 10 to 100 /xg of the pure nonradioactive substance 
with which the radioactive compound is thought to be iden- 
tical. The new chromatogram is then developed. A radio- 
autograph is prepared to locate the radioactive substance, 
after which the paper is sprayed with a chemical spray (for 
example, ninhydrin for amino acids), which produces a color 
where the carrier compound is located on the paper. Super- 
position of the paper chromatogram and the radioautograph 
(x-ray film) will show an exact coincidence between chem- 
ically developed color on the paper and the black spot on 
the film, provided the two substances are identical. 

Once the identity of the radioactive compounds formed 
during a short period of photosynthesis had been established, 
experiments were performed under a variety of conditions 
and times of exposure of the algae to radiocarbon. 

The radioautogram from the experiment with Chlorella 
described above is shown in Figure 3. Even after only 10 
seconds of exposure to C^^, a dozen or more compounds are 


found. Some of these (the sugar phosphates) are not sepa- 
rated from each other by the first chromatography and must 
be subjected to further analysis. When the sugar mono- 
phosphates are hydrolyzed to remove the phosphate groups 
and rechromatographed, separate spots are found of triose 
(dihydroxyacetone), tetrose, pentoses (ribulose, xylulose, and 
ribose), hexoses (glucose and fructose), and heptose (sedo- 
heptulose). The radioactive sugar diphosphates area gives 
free ribulose, fructose, glucose, and sedoheptulose. 

After periods of photosynthesis with C^* of less than 5 
seconds, 3-phosphoglyceric acid (PGA) was found to be the 
predominant radioactive product. Chemical degradation of 
this compound showed that the radioactivity first appears in 
the carboxyl carbon (14). Later kinetic studies showed that 
the rate of incorporation of C^^ into PGA at very short 
times was much greater than the rate of labeling of any 
other compound (18,1). Therefore, it was concluded that 
PGA is the first stable product of carbon dioxide fixation 
during photosynthesis, and, furthermore, that carbon dioxide 
first enters the carboxyl group of PGA, presumably via a 
carboxylation reaction. 

Further reactions in the photosynthetic sequence were 







Figure 3. Radioautograph of two-dimensional paper chromato- 
gram. Alcoholic extract of Chlorella pyrenoidosa after 10 seconds 
photosynthesis with Ci'*02. 


suggested by the already known pathways of the glycolytic 
breakdown of sugars, which lead to PGA as an intermediate. 
Since the sugar phosphates are important early products of 
carbon reduction in photosynthesis, it was proposed that 
they are formed from PGA by a reversal of the glycolytic 
pathway. Degradation of the radioactive hexoses from short 
experiments showed that they were labeled in the two center 
carbon atoms (numbers 3 and 4) just as one would expect if 
2 molecules of carboxyl-labeled PGA were first reduced to 
triose and then linked together by the two labeled carbon 
atoms to give hexose (Figure 4). 

The hexose and triose phosphates may be converted by 
aldolase or transaldolase and transketolase enzymes to pen- 
tose and heptose phosphates (Figure 2 and Table 2). Deg- 
radation of these sugars and comparison of the labeling pat- 
terns within the molecules showed that this conversion did 
occur, and in such a way that 5 molecules of triose phos- 

and pentose 

Hexose phosphotes 

Triose phosphate 














3- PGA 

2- PGA 




















t . 





C = 









Malic ocid 


Figure 4. Labeling of compounds with C 
during early steps in carbon dioxide reduction 
during photosynthesis with C^*02- 


phate were ultimately converted to 3 molecules of pentose 

Other known metabolic pathways leading from PGA 
(Figure 4) give rise first to phosphoenolpyruvic acid (PEP A), 
which then may undergo further transformations, including 
the following: (1) it may be carboxylated and transaminated 
to give aspartic acid, (2) it may be carboxylated and reduced 
to give malic acid, or (3) it may be dephosphorylated and 
transaminated to give alanine. All these compounds are 
labeled after short exposures of the algae to HC"0.s~ in the 

The enzyme system of plants, which during respiration 
brings about the oxidation of triose phosphate to PGA in 
the glycolytic pathway, was known to produce ATP and 
TPNH (or DPNH). If PGA is to be reduced to triose phos- 
phate during photosynthesis, it follows that ATP and TPNH 
must be supplied. We have already seen that these two co- 
factors, and possibly others, are produced as a consequence 
of the light reaction and the splitting of water. It might be 
expected that, if the light were turned off from plants photo- 
synthesizing in ordinary carbon dioxide at precisely the same 
time that C^^02 is introduced, PGA would no longer be 
reduced to sugar phosphates but would still be formed (if 
no light-produced cofactors are required for the carboxyla- 
tion reaction). Moreover, the PGA would still be used in 
other reactions not requiring these cofactors. In Figure 5, 
the radioautograph from just such an experiment, this pre- 
diction proves to be correct. Labeled PGA is still formed 
by the algae from C^^02 during 20 seconds in the dark, but 
only a very little of the PGA is reduced to sugar phosphates. 
At the same time, a large amount of alanine is formed from 
PGA via PEPA in reactions that do not require ATP. 
The trace of labeled sugar phosphates that does appear may 
be due to the residual ATP, or some unknown cofactor, 
which was formed while the light was on but which had 
not yet been used up when the C^^02 was introduced. The 
formation of malic acid and of alanine and aspartic acid 


in the dark indicates the presence of some reducing cofactors, 
either remaining from the light or derived from some other 
metabolic reaction. 

Before we discuss the evidence for the remainder of the 
carbon reduction cycle, we must describe another type of 
experiment with C^^Oa and photosynthesizing algae. In these 
experiments, algae are first permitted to photosynthesize for 
20 minutes or more in the presence of a constant supply 
of C^*02. During this time environmental conditions are 
maintained nearly constant (temperature, CO2 pressure, 
light intensity, etc.). After about 10 minutes of exposure to 
C^*02, so much radiocarbon has passed through the various 
biochemical intermediate compounds on its way to end 
products that each carbon atom of each intermediate com- 
pound contains, on the average, the same percentage of C^^ 
atoms as the CO2 being absorbed. In other words, the specific 
radioactivities of all the carbon atoms of all the early inter- 
mediates are the same as the specific radioactivity of the 
entering radiocarbon, which can be measured. 

At this point samples of the algae are removed without 
disturbing the rest of the algae, and these samples are killed 
and subsequently analyzed by the methods described. The 






20 SEC DARK C'^Oj FIXATION ^0 ^''^ 


Figure 5. Radioautograph of chromatogram of products of 
20 seconds C^^Oa fixation by Chlorella pyrenoidosa in the dark 
following a period of photosynthesis. 


total radioactivity of each intermediate is measured, and, 
when this is divided by the known specific radioactivity of 
the entering CO2, the total number of carbon atoms of each 
intermediate compound in the sample can be calculated. 
Thus the number of moles per unit volume of algae of the 
various intermediates of the actively photosynthesizing sys- 
tem may be determined. This number of moles per unit 
volume of plant material is an average concentration, since 
the distribution of molecules in such a heterogeneous system 
is not homogeneous. 

This determination of the concentrations of intermedi- 
ates in vivo is an extremely valuable tool which has many 
uses, but let us proceed for the moment with one particular 
application. Having taken a sample of algae for later de- 
termination of the concentrations of compounds, the experi- 
menter turns off the light and proceeds to take a series of 
samples of the algae as rapidly as possible, which is about 
every 3 seconds. When the concentrations of compounds in 
these samples are determined, any changes resulting from 
turning off the light will be revealed. The two most striking 
changes are found to be in the concentration of PGA, which 
increases rapidly, and in the concentration of one particular 
compound, ribulose diphosphate, which drops rapidly to 
zero (16,20). 

The increase in PGA on turning off the light is expected. 
The cofactors, derived from the light reaction, are necessary 
for the reduction of PGA. The rapid drop in ribulose diphos- 
phate, taken together with the fact that other sugar phos- 
phates initially do not drop rapidly in concentration, indi- 
cates that the formation of ribulose diphosphate from other 
sugar phosphates requires a light-formed cofactor. This con- 
clusion agrees with the fact that the known enzyme, which 
converts ribulose-5-phosphate to ribulose- 1,5-diphosphate 
(RuDP), requires ATP (Table 2). The drop in ribulose di- 
phosphate, alone among the sugar phosphates, means that it 
is being used up by a reaction that does not require light. 

Ribulose diphosphate, then, is used up by some reaction 


that proceeds in the dark, and PGA continues to be formed 
in the dark. Could the carboxylation of ribulose diphosphate 
to form PGA be the first step in carbon dioxide reduction? 
To answer this question, an experiment similar to the one 
just described was performed. This time, however, instead 
of turning off the light, the light was left on, and carbon 
dioxide was suddenly removed (19). The result of this ex- 
periment was that the concentration of ribulose diphosphate 
now rose rapidly while PGA dropped rapidly. Thus the car- 
boxylation of RuDP to give PGA was confirmed. 


^^ carboxylation 
'^^ reactions 

Thus far we have mentioned two carboxylation reactions 
in photosynthesis: carboxylation of RuDP (the carbon re- 
duction cycle) and carboxylation of PEPA. When algae 
have been allowed to photosynthesize for less than a minute, 
virtually all the radioactivity found on the chromatogram 
prepared from the algae is located in compounds apparently 
derived from these two reactions. There still remained the 
possibility that other carboxylation reactions might occur 
which would involve intermediate compounds too unstable 
or too volatile to be seen on the chromatograms. These pos- 
sibilities were tested by making a quantitative comparison 
between the rate of uptake of C^^02 from the medium and 
the rate of appearance of C^* in compounds on the chromato- 
grams (3). 

For these experiments, the algae were kept, as close as 
possible, in steady-state growth in the experimental vessel. 
Light, temperature, pH, and supply of inorganic nutrients 
were kept constant. Gas was circulated through the algae 
suspension in a closed system by means of a pump. Levels of 
CO2, O2, and, when present, C^^Oo, were continuously 
measured and recorded. From the known gas volumes of the 


system and the recorded rates of changes in gas tensions, 
we calculated the total change in these gases as a function 
of time. Then we added 0^*02 to the system and took sam- 
ples of algae every few seconds for the first few minutes and 
then less frequently up to an hour. Each sample of algae was 
killed immediately and a portion analyzed as described 
earlier. A part of each sample was reserved and was dried on 
a planchet to determine the rate of appearance of C^* in all 
stable nonvolatile compounds. This rate proved to be the 
same as the externally measured rate of uptake of CO2 and 
C^^ between 20 and 60 seconds after the addition of C^^. If 
unstable or volatile intermediates do precede these stable 
compounds, they are equivalent in micromoles of carbon 
to no more than 5 seconds photosynthetic fixation, according 
to the shape of the fixation curve during the first 20 seconds. 
We analyzed each sample by paper chromatography 
and determined the radioactivity in each compound in each 
sample. On the basis of the externally measured uptake rates, 
at least 85 per cent of the carbon was found to be incorpo- 
rated into individual compounds on the paper chromato- 
grams during the first 40 seconds. At least 70 per cent of the 
total carbon uptake rate could be accounted for by the ap- 
pearance of C^** in compounds apparently derived from the 
RuDP carboxylation reaction of the carbon reduction cycle 
via the pathways shown in Figure 2. Another 5 per cent or 
more was found to be incorporated via C1-C3 carboxylation. 
About 5 per cent was found in unidentified compounds or 
in glutamic acid, whose photosynthetic pathway is not defi- 
nitely known. Of the 15 per cent not accounted for, some 
may be in nonextractable polysaccharides, whose sugar phos- 
phate precursors become labeled very quickly. More of the 
unaccounted-for radiocarbon is undoubtedly in a large num- 
ber of unmeasured compounds on the chromatograms. Each 
of these compounds contains by itself too little C^"* to be 
readily determined. In any event, it is clear that the known 
fixation pathways are the only quantitatively important 


ones unless there are unknown pathways utilizing the same 
intermediate compounds. 

A kinetic analysis of the appearance of C^'' in PGA 
and RuDP in this experiment indicated that the carboxy- 
lation reaction results in the formation of only one free mole- 
cule of PGA per molecule of CO2 entering the cycle. The 
kinetic analysis cannot say what the other three-carbon 
fragment would be. It might be merely a molecule of PGA 
bound in some way so that its labeling remains distinct from 
that of the PGA from the other half of the six-carbon addi- 
tion product. The only other compounds that seem to satisfy 
the kinetic requirements and that could readily result from 
the splitting of the six-carbon addition product are the triose 
phosphates. The formation of a molecule of triose phosphate 
in this way would require a reductive split of the addition 
product, as indicated by the dashed line in Figure 2. 

That such a pathway differing from the in vitro reaction 
may exist seems entirely reasonable, since the enzymes of the 
carbon reduction cycle appear to be closely associated with 
the molecular structures in which the TPNH is formed in 
the chloroplast (21). In the intact plant the carboxylation 
enzyme, as well as the enzyme responsible for the splitting of 
the product and the enzyme that brings about the reduction 
of TPN+ to TPNH, might be part of a structurally organ- 
ized system. In fact, if a reductive scission does occur, the 
reducing agent could be a substance formed from the oxida- 
tion of water and preceding TPNH in the electron transport 
chain. This substance might never be available in sufficient 
concentration to be a factor in in vitro systems in which 
carboxydismutase is coupled with isolated or broken chloro- 
plasts. Such an explanation of the experimentally observed 
kinetic result is purely hypothetical. We mention it to focus 
attention on the possibility that a given biosynthetic pathway 
may follow a different course in an intact cell than that which 
would be predicted on the basis of studies with fragmented 
cells or enzymes alone. 


In higher plants much of the product of photosynthesis 
must be transported to a nonphotosynthetic part of the plant. 
This requires that higher proportions of easily transported 
molecules such as sucrose are formed (4). In all higher plants 
that have been studied, however, there is appreciable direct 
photosynthesis of amino acids and fats, not just carbohydrates. 


^g^ Balance among 
"^!^ synthetic pathways 

We have seen that in each complete turn of the carbon- 
reduction cycle 3 molecules of RuDP (15 carbon atoms) are 
carboxylated by 3 molecules of CO2 to give 6 three-carbon 
compounds (18 carbon atoms). Thus there is a net gain of 3 
reduced carbon atoms. These atoms are withdrawn from the 
cycle for further synthesis. They may be withdrawn from 
the cycle as PGA or as any of the sugar phosphates in the 
cycle. Before the photosynthetic reactions had been mapped, 
it was commonly believed that photosynthesis leads first to 
carbohydrates only and that these carbohydrates are then 
converted via nonphotosynthetic reactions to other com- 
pounds such as amino acids and fatty acids. We now know 
that pathways leading from the carbon reduction cycle to 
amino acids and fatty acids and other substances can be just 
as important quantitatively as those leading to carbohydrates. 
This is particularly true in a unicellular algae, as exempli- 
fied by Chlorella pyrenoidosa, where under some conditions 
less than half of the assimilated carbon is directly converted 
into carbohydrate. This carbohydrate synthesis draws its 
carbon from the cycle in the form of sugar phosphates. Con- 
sequently, more than half of the carbon drained from the 


carbon reduction cycle as PGA or sugar phosphates may be 
used in fat and protein synthesis. 

It is interesting to consider an extreme case in which 
all the carbon assimilated by the carbon reduction cycle 
would be withdrawn from the cycle as PGA, converted to 
PEPA, and then carboxylated to give four-carbon compounds. 
In this case, 75 per cent of the assimilated carbon would 
enter the photosynthetic pathways via the carbon reduction 
cycle, while the remaining 25 per cent would enter via the 
carboxylation of PEPA. 

With normal conditions of steady-state growth under 
high light intensity, the ratios of various fixation pathways 
must be determined to a large extent by the requirements 
of the plant for the small molecules from which the protein, 
carbohydrate, fat, and other substances of the plant are syn- 


^<p vs. other forms 
'^^ of biosynthesis 

Biosyntheic reactions in plants cannot be classified as 
photosynthetic or nonphotosynthetic on the basis of direct 
photochemical action because all reactions in the synthetic 
pathways are probably "dark" reactions. However, we can 
make a classification on the basis of the immediate source of 
the required cofactors. The conversion of light energy results 
in the formation of ATP and TPNH and perhaps other un- 
known cofactors. When these cofactors are formed by the 
light reaction and are used to bring about the synthesis of 
carbon compounds, we may consider the reactions to be 
photosynthetic. Also included in this category would be 
preliminary and intermediate steps such as hydrations, con- 
densations, and carboxylations. 

It may well be that all photosynthetic reactions, as just 
defined, occur in the chloroplasts while the light is on. 
If this is true, reactions outside the chloroplast would derive 
their energy from substrate carbon compounds which diffuse 
from the chloroplast to the extrachloroplastic spaces of the 
cell. Such an interpretation is suggested by the report by 
Tolbert (22), who found that chloroplasts isolated from Swiss 
chard, when allowed to photosynthesize with HC^^Oa", ex- 


creted mainly glycolic acid into the medium. Phosphate 
esters, of importance to the carbon reduction cycle, were 
retained in the chloroplasts. Isolated chloroplasts have a 
carbon metabolism that is much more limited than photo- 
synthesis in intact cells. This is probably due to loss of 
enzymic activity by chloroplasts during the isolation process. 
In all probability the carbon compounds excreted by intact 
chloroplasts in vivo include substances other than glycolate. 
There is more than a semantic reason for making a dis- 
tinction between photosynthetic and nonphotosynthetic 
pathways. The environment of the photosynthetic metab- 
olism is unique. There is an abundance of the reduced and 
energetic form of the coenzymes. Hence synthetic pathways 
do not require energy derived from degradative reactions 
such as decarboxylations and oxidations. For example, a 
well-known biosynthetic pathway leading to glutamic acid 
from acetate includes oxidative and decarboxylation steps. 
Such a pathway is to be expected in a nonphotosynthetic 
system, where degradation of part of the substrate is the 
only means of obtaining the energy and reducing power for 
synthetic reactions. In a photosynthetic system one might 
expect instead a pathway involving only condensations, re 
ductions, and carboxylations. We cannot say that this differ- 
ence in type of reaction will always be borne out by the 
actual mechanisms when they are known. This proposed dif- 
ference in reaction type may be a useful working hypothesis 
to those who attempt to map photosynthetic pathways from 
experimental data. 



Amino acid 

"^^ synthesis 

Among the first compounds found to be labeled by pho- 
tosynthesis of C^*02 in algae were alanine, aspartic acid, and 
several other amino acids (11). These compounds were 
slowly labeled even in the dark when algae were exposed 
to C^^02. They and malic acid were much more rapidly 
labeled if the algae were photosynthesizing, or had been 
photosynthesizing, just prior to the moment of addition of 
C^^02. We recognized that these amino acids were therefore 
products of photosynthetic reduction of CO2, even though 
they could also become labeled by reversible respiratory re- 
actions. Accelerated incorporation of C^^ into amino acids 
in higher plants during photosynthesis has been noted in 
this laboratory (23,24) and in many others (25-28). Nichi- 
porovich (25) has presented and reviewed evidence that syn- 
thesis of proteins in the chloroplasts of higher plants is 
greatly accelerated during photosynthesis. This accelerated 
protein synthesis appears to occur directly from the inter- 
mediates of photosynthetic carbon reduction, since the pro- 
teins were labeled when C^^02 was used but not when C^'*- 
labeled carbohydrate was administered. Photosynthetically 
accelerated synthesis of protein containing N^^ was also ob- 


served when N^^H4+ was administered. Sissakian (29) has 
reviewed evidence that protein can be synthesized in isolated 
chloroplasts from nonprotein nitrogen, including peptides. 

In experiments in this laboratory (30) it recently has 
been possible to measure the proportion of the total carbon 
fixed by Chlorella pyrenoidosa, which is directly incorporated 
into certain key amino acids. These experiments show that, 
during steady-state photosynthesis in bright light with an 
adequate supply of inorganic nutrients, the synthesis of these 
amino acids can account for 60 per cent of all the carbon 
fixed by the algae and 30 per cent of the uptake of NH4+, 
which is also measured. If the light is turned off, the NH4 + 
uptake and C^* fixation into amino acids are both accelerated 
for about 10 minutes and then drop to a very small fraction 
of the rates in the light. Finally, these experiments indicate 
clearly that in Chlorella pyrenoidosa there are at least two 
pools of alanine, glutamic acid, aspartic acid, and serine, and 
probably other amino acids as well. One of these pools, 
especially in the cases of alanine and aspartic acid, is labeled 
extremely rapidly after the introduction of C^*02 to the 
algae. So rapidly are these compounds labeled, in fact, that 
the site of their synthesis must be freely accessible to their 
photosynthetically formed precursors, namely, phosphoenol- 
pyruvic acid and PGA (see Figure 1). The studies of Tolbert 
(22) and Moses et al. (31) indicate that the photosynthetic 
pools are isolated from the extrachloroplastic region. We 
conclude, therefore, that in Chlorella the more rapidly 
labeled pools of amino acids are located at the site of photo- 
synthetic carbon reduction, probably in the chloroplast. 

The rates of flow of carbon through these pools of amino 
acids as determined from kinetic labeling data with Chlorella 
in a typical experiment are shown in Table 1. 

The amino acids shown in Table 1 are those most prom- 
inently labeled with C^^ during a few minutes of photosyn- 
thesis. In addition, a nuinber of other amino acids become 
labeled as time passes. The rates of labeling seem to indicate 
that the carbon skeletons of these other acids are probably 


Table 1 
Rates of Flow of Carbon through Active Pools of Amino Acids 

Calculated rate Equiv. NH4"'" 
of synthesis R, uptake, Mmoles 
Compound jumoles of carbon of NH4"'" 







Aspartic acid 



Glutamic acid 


















Externally measured 




% of total through these pools 



* Not included in totals. 


Figures are for carbamyl carbon only. 

derived, for the most part, from the listed amino acids. How- 
ever, the aromatic rings of the amino acids are synthesized by 
another pathway. 

In Table 1 we compare the rates of synthesis of carbon 
skeletons that have been measured with the rate of uptake 
of NH4 + . The rate of synthesis of any given amino acid 
does not necessarily represent the rate of incorporation of 
inorganic nitrogen into that amino acid, since it could be 
formed by transamination from another amino acid. How- 
ever, the total of the rates of synthesis of all "primary" amino 
acids should account for the major fraction of the rate of 
uptake of ammonia. By "primary" amino acids we mean 
those amino acids whose carbon skeletons are not synthesized 
from some other amino acid. Alanine, serine, and aspartic 
acid are clearly primary amino acids, since their rates of 
labeling reach a maximum as soon as the intermediates in 


the carbon reduction cycle are saturated (about 3 minutes in 
this experiment) and long before they themselves, or any 
other amino acids, are saturated with radiocarbon (30). Prob- 
ably glutamic acid is a primary amino acid also, but kinetic 
data alone cannot prove this at the moment. Glutamine is 
generally supposed to arise from glutamic acid, but there is 
some evidence to indicate that it may arise as a primary 
amino acid amide (32,30). 

In any event, the rates of synthesis of alanine, serine, and 
aspartic acid in reservoirs we believe to be closely associated 
with the chloroplasts in Chlorella are great enough to permit 
the following conclusions. 

1 . An appreciable fraction of the carbon assimilated dur- 
ing photosynthesis in Chlorella is used directly in the syn- 
thesis of amino acids without the intermediacy of sugars or 
any other class of compounds except acid phosphates and 
carboxylic acids. 

2. Since this amino acid synthesis accounts for a major 
portion of the inorganic nitrogen uptake, these amino acids 
must be used to a large extent in protein synthesis. However, 
some important amino acids (i.e., glycine) are so slowly 
labeled that they probably do not supply a major part of 
the carbon for protein synthesis. Instead, the carbon skeletons 
corresponding to these amino acids must be incorporated into 
protein in some form other than as the free amino acid. 

Before considering synthetic routes to specific amino 
acids, we wish to reiterate our belief that photosynthetic re- 
actions need not follow the same course as the better-known 
synthetic reactions of other nonphotosynthetic organisms. 
Also note that few if any enzymes involved in amino acid 
synthesis have ever been isolated from chloroplasts. Thus 
we are forced to suggest new and untested hypothetical paths. 
Our guiding principles will be that chemical potential should 
be used to drive the reactions rapidly in the forward direction 
and that loss of carbon or reduction level should be avoided 
wherever possible. 


In Figure 6 are shown hypothetical pathways leading 
from PGA to alanine, serine, aspartic acid, and malic acid. 
These pathways differ somewhat from known enzymatic 
pathways, in that, in each step leading to the amino acid, 
ammonia reacts with a phosphoric acid ester. 

The rapid incorporation of inorganic nitrogen into 
organic compounds would be brought about by the large 
negative free-energy change associated with each of these re- 
actions. Thus these reactions, and not the reductive amina- 
tion of ketoglutaric acid alone, would account for a major 
portion of ammonia incorporation during photosynthesis. 


Reduction Cycle 





I I 

HC-0® '^"^ HC-NH2 + HO (?) 

CO2" CO2" 






" r^ NH4+ I ^ 

C-0(P) ». HC-NH2 + HO (P) 




















TPNH HC-NH2+H0(£) 



Figure 6. Hypothetical pathways of photosynthesis of alanine, 
aspartic acid, serine, and malic acid. 


This seems entirely reasonable when one considers that PGA 
is both the immediate precursor in these reactions and the 
primary product of carbon reduction during photosynthesis. 
These amino acids could then supply ammonia via transami- 
nase reactions for the synthesis of many other amino acids. 
Holm-Hansen et al. (33) have demonstrated the presence of 
a transaminase activity in spinach chloroplasts, which is very 
effective in the transfer of amino groups from unlabeled ala- 
nine to C^^-labeled pyruvic acid. 

The three-carbon precursors to these amino acids are in 
rapid equilibrium with PGA. PEPA becomes C^^-saturated 
during photosynthesis in C^^02 in Chlorella almost as soon 
as PGA itself. The proposed phosphoenoloxalacetate prob- 
ably does not exist except in enzyme complexes. Thus, by 
the time the PGA is C^*-saturated, these amino acids are 
being labeled as rapidly as if they were formed directly from 

It has been suggested that glutamic acid is formed dur- 
ing photosynthesis by a carboxylation of y-aminobutyric acid 
(34). Judging by our studies with Chlorella pyrenoidosa dur- 
ing steady-state photosynthesis with C^''02, this reaction ap- 
parently does not constitute a source of glutamate, since y- 
aminobutyric acid does not become labeled, even by the time 
the glutamic acid is 50 per cent saturated with C^* and long 
after the rate of labeling of glutamic acid has passed its maxi- 
mum. Clearly, a compound cannot be a precursor in a steady- 
state system unless it is itself continuously regenerated. If 
the reaction does occur at all, the glutamic acid so formed 
could only be a shuttle for CO2, regenerating unlabeled y- 
aminobutyric acid. Even so, such a carboxylation reaction 
does not account for more than about I per cent of the car- 
bon fixed in our studies of steady-state CO2 fixation by Chlo- 

One possible route from PGA to glutamic acid would 
begin with conversion of PGA to PEPA, followed by car- 
boxylation of PEPA to give oxalacetic acid. Condensation 
of oxalacetic acid with acetyl CoA would give citric acid, 


thence aconitic acid, thence isocitric acid. Proceeding along 
the Kreb's cycle, the next two steps are oxidation to oxalo- 
succinic acid, followed by oxidation and decarboxylation 
to give a-ketoglutaric acid. Finally, the reductive amination 
would give glutamic acid. This pathway may be followed in 
Chlorella pyrenoidosa in the synthesis of glutamic acid, par- 
ticularly when the light is turned off. We suspect that it is 
not the principal pathway during photosynthesis for two rea- 
sons, one experimental and one theoretical. Experimentally, 
the rates of labeling of the intermediate compounds such as 
citric acid and ketoglutaric acid are too slow to permit them 
to serve as precursors to the more rapidly labeled reservoir 
of glutamic acid. Theoretically, the pathway is objectionable 
to us as a photosynthetic route because it involves two oxida- 
tions and a decarboxylation. 

How else might glutamic acid be formed during photo- 
synthesis? The availability of three-carbon and two-carbon 
compounds suggests the possibility of a simple condensation. 
Barker and co-workers (35-37) found an enzymic pathway in 
certain microorganisms leading from glutamic acid to py- 
ruvic acid and acetate via citramalate, mesaconic acid, and 
/3-methylaspartate. The reverse of this pathway might operate 
during photosynthesis also. However, we have been unable 
so far to find significant amounts of radiocarbon in either /?- 
methylaspartic acid or mesaconic acid in Chlorella which 
were synthesizing glutamic acid from C^'*02. Moreover, a gen- 
eral energy-conserving principle would suggest that PEPA 
and not free pyruvic acid should be the three-carbon com- 
pound that combines with the two-carbon fragment. As we 
shall see in the discussion for the synthesis of aromatic rings, 
it has been proposed that PEPA can condense with an alde- 
hyde, erythrose phosphate, to give (eventually) phosphoshi- 
kimic acid (38). Perhaps a similar reaction between PEPA 
and glyoxylic acid could lead to a product such as y-hydroxy- 
glutamic acid, which could be subsequently converted to 
glutamic acid. Dekker (39) has reported the presence of an 
enzyme in rat liver that converts y-hydroxyglutamic acid to 


glyoxylate and another product, which may be alanine. The 
presence of y-hydroxyglutamic acid in green leaves has been 
reported by Virtanen and Hietala (40). The dehydration and 
reduction of y-hydroxyglutamic acid to give glutamic acid 
would be common types of biochemical reactions, analogous 
to the formation of succinic acid from malic acid. However, 
we have at present no experimental evidence for such a path- 

Threonine does not become labeled as rapidly as the 
amino acids so far discussed, and it may well be secondary in 
origin. That is, it may be an example of conversion of pri- 
mary amino acids (aspartic acid, alanine, serine, and glutamic 
acid) to other amino acids of their respective families, a proc- 
ess that presumably occurs in photosynthesis. 

The small amount of labeled glycine formed during 
steady-state photosynthesis may come from either serine or 
glyoxylic acid. 


'^f' Carboxylic acids 

Malic and fumaric acids 

Malic acid and fumaric acid are rapidly labeled during 
steady-state photosynthesis with €^^^02. These acids are 
probably formed by reduction of the product of carboxyla- 
tion of PEPA. In the steady-state experiment that yielded the 
results shown in Table 1, about 5 per cent of the C^^ uptake 
rate could be accounted for in the labeling of these two 
acids. In that experiment very little of the radioactivity finds 
its way into succinic acid. It would thus appear that, if malic 
and fumaric acids are labeled by reductive carboxylation of 
PEPA, either (1) the reaction sequence is highly reversible, 
leading to exchange labeling, or (2) the malic and succinic 
acids are converted to other compounds by as-yet-undeter- 
mined paths. 

The probability of labeling via exchange (1) may be 
answered by a thermodynamic argument. Under the condi- 
tions existing in the chloroplast during photosynthesis, the 
actual free energy change accompanying the conversion of 
PEPA, CO2, TPNH, and either ADP or IDP to malic acid, 


TPN + , and ATP or IT? is probably at least -7 kcal. The 
ratio of the forward reaction to the back reaction, given by 

DT-i /forward rate' 
V back rate 

/back rate plus net rate\ 
\ back rate J 

would thus be 10^ or greater. Since the rate of labeling of 
malic acid is measurable and gives the net rate by a simple 
calculation, the back reaction, and hence the exchange label- 
ing, can be shown to be of negligible importance. 

This type of calculation is of considerable importance in 
in vivo steady-state kinetic calculations. Another example is 
the conversion of malic acid to fumaric acid. In this case, the 
actual free energy change is small; the two acids are essen- 
tially in equilibrium with respect to C^*-labeling. Thus the 
sum of the pools of the two acids can be treated from a 
labeling standpoint as a single entity. 

In any event, if malic acid is not labeled by exchange 
and is not converted to succinic acid yet is being formed at 
a rapid rate under steady-state conditions, it must undergo 
some as-yet-unknown conversion. One possibility might be 
that it is split to give glyoxylic acid and free acetate. The 
actual free energy change for such a reaction under steady- 
state conditions would be negative, whereas the reaction to 
give glyoxylic acid and acetyl Co A would probably be posi- 
tive and the latter reaction would not occur. Acetate could 
be converted to acetyl phosphate with ATP and then to 
acetyl CoA. The acetyl CoA thus formed could be used in 
fatty acid synthesis and other biosynthetic reactions. The 
glyoxylic acid could be used in the synthesis of glycolic acid, 
glycine, and possibly, as suggested in the previous section, 
glutamic acid. 

The synthesis of labeled malic acid could occur via 
condensation of glyoxylate with acetyl CoA, provided there 
is some other route for the labeling of these two-carbon 
acids (such as are suggested later). It is quite likely that malic 


acid is so synthesized in the cytoplasm, outside the chloro- 
plasts. Within the chloroplasts, however, the appearance of 
C^^ in malic acid in the very shortest exposures to C^*02 and 
in the pre-illumination experiments (see Figure 5) indicate 
that it is, in part at least, a product of Ci-Cs carboxylation 
and reduction. 

Glycolic acid, acetic acid, and glyoxylic acid 

Even if acetate and glyoxylate are formed from malic 
acid, there are probably other more important synthetic 
routes from the carbon reduction cycle to these compounds. 
Benson and Calvin (41) found that barley seedlings sub- 
jected to 30 seconds photosynthesis with C^^02, followed by 
2 minutes light without CO2, formed large amounts of C"- 
labeled glycolic acid. Calvin et al. (14) and Schou et al. (42) 
degraded glycolic acid and phosphoglyceric acid obtained 
from barley leaves and from Scenedesmus that had photo- 
synthesized for a few seconds in the presence of C^*02 or 
HC^^Os". The alpha and beta carbon atoms of PGA were 
found to be always about equal to each other in radioactivity 
and always less than the carboxyl carbon until such time 
(1 to 5 minutes) as all three carbon atoms were completely 
labeled. The two carbon atoms of glycolic acid were always 
about equal to each other in labeling. When C^*H20H 
— COOH was administered to the unicellular algae Scenedes- 
mus during 10 minutes photosynthesis with V2 per cent CO2 
in air or N2, a pattern of photosynthetic intermediates was 
found similar to that obtained during photosynthesis with 
C^*02. Moreover, upon degradation of the PGA we found that 
less than 10 per cent of the radioactivity was in the carboxyl 
carbon. Clearly, glycolic acid is incorporated for the most part 
into normal intermediates of the carbon reduction cycle with- 
out preliminary conversion to CO2, since so little C^* was 
found in the carboxyl carbon of PGA. However, alpha and 
beta carbon atoms of the PGA were found to be equally 
labeled. Thus the pathway from glycolic acid to the alpha and 


beta carbon atoms of PGA involves a randomization of the 
label. This could mean that along this pathway there is a 
symmetrical intermediate or that an intermediate is in rapid 
reversible equilibrium with a symmetrical compound (see 
Figure 7). 

When Wilson and Calvin (19) studied the effect of CO2 
depletion following a period of photosynthesis with C^^02 
by algae, they found that the lowering of CO2 pressure re- 
sulted in a great increase in the amount of labeled glycolic 
acid. This increase in labeled glycolic acid was sustained for 
at least 10 minutes. Upon application of 1 per cent CO2 
again, the level of labeled glycolic acid declined. 

Tolbert (22) found that glycolic acid formation from 
C^^02 during 10 minutes photosynthesis in leaves of Sedum 
alboresum is much higher at very low CO2 pressure than at 
high CO2 pressure. As mentioned earlier, he also found that 
glycolic acid is the predominant labeled compound excreted 
into the medium by chloroplasts from Swiss chard photo- 
synthesizing in the presence of HC^^Os-. He had shown 
earlier (43) that glycolic labeled with C^* is excreted into 
the medium by Chlorella photosynthesizing in C^^02. He 
suggested that glycolate may function in ion balance with 
HCO3- between cells and their medium or between chloro- 
plasts and other cell compartments. He also proposed that 
glycolate might be a carrier of "carbohydrate reserves" from 
the chloroplasts to the cytoplasm. 

Moses and Calvin (44) exposed photosynthesizing Chlo- 
rella pyrenoidosa to tritium-labeled water for various periods 
from 5 seconds to 3 minutes. Analysis was made by the usual 
extraction, two-dimensional paper chromatography, and 
radioautography. The greatest darkening of the film by far 
occurred where it was in contact with the glycolic acid area 
of the chromatogram. This result, which we shall discuss 
later, seems to agree with Tolbert's suggestion that the gly- 
colic acid acts as a carrier of hydrogen. 

During normal photosynthesis (Figure 2), two-carbon 
moieties (carbon atoms number 1 and 2 from a keto sugar 


phosphate) are transferred during a reaction similar to that 
catalyzed by transketolase (45,46) to an aldo-sugar phosphate, 
producing a new ketose phosphate, two carbon atoms longer 
than the starting aldose. Other enzymes have been found in 
nonphotosynthetic organisms which convert the carbon 
atoms number 1 and 2 of a ketose phosphate to acetyl phos- 
phate, leaving the remainder of the sugar as an aldose phos- 
phate. One of these is phosphoketolase (47), which is specific 
for xylulose-5-phosphate, while another is fructose-6-phos- 
phate ketolase (48), which can act on either fructose-6-phos- 
phate or xylulose-5-phosphate. These enzymes require thia- 
mine pyrophosphate, inorganic phosphate and, in some cases, 
Mg+ + . Stimulation by Mn++ or Ca++ in place of Mg+ + 
could sometimes be observed, whereas levels of Mn above 
10~^ were inhibitory. 

Breslow has proposed a mechanism for the role of thi- 
amine pyrophosphate in these reactions (49,50). In his mech- 
anism, some of which forms the basis for part of Figure 7, 
the hydrogen at position 2 of the thiazole ring is an active 
hydrogen which can dissociate from the acidic carbon at that 
position to give a carbanion. This carbanion adds to the Qar- 
bonyl carbon of the ketose (somewhat analogous to cyan- 
hydrin addition). The bond between carbons 2 and 3 of the 
ketose breaks, with the electron pair going to the reduction 
of carbon 2 of the ketose, to give a glycolaldehyde-thiamine 
pyrophosphate. The remainder of the sugar becomes an 
aldose. Reversal of this reaction path, with a different aldose, 
completes the transketolase reaction. 

Alternatively, glycolaldehyde-thiamine pyrophosphate 
may eliminate the elements of water (OH~ from the beta 
carbon and H+ from the alpha carbon of the glycolaldehyde 
moiety) to give the enol form and thence the keto form of 
acetyl-ThPP. This compound can undergo phosphoroclastic 
cleavage to give acetyl phosphate and thiamine pyrophos- 
phate (ThPP). 

The mechanisms find support in the demonstration by 
Breslow that the hydrogen atom on the number 2 position 














^\ 4[H] 




/ \ 


Figure 7. Pathways from carbon reduction cycle to acetyl phos- 
phate and glycolic acid. For details of the carbon reduction cycle, 
see Figure 2. 

of the thiazole ring does exchange rapidly in D2O (49). In 
support of an analogous mechanism for the role of ThPP 
in the oxidation of pyruvate, Krampitz and co-workers (51, 
52) synthesized the postulated intermediate, an acetaldehyde- 
ThPP compound with the acetaldehyde bonded to the num- 
ber 2 carbon atom of the thiazole ring as an alpha hydroxy- 
ethyl group. This compound was found to be active in the 
reactivation of carboxylase and also to be capable of non- 
enzymatic reaction with acetaldehyde to give acetoin. The 
postulated mechanism for the oxidation of pyruvic acid thus 
begins with a reaction between pyruvate and ThPP to give 
addition of the carbonyl carbon to the thiazole-ring-position 
number 2. Concurrently or immediately following this addi- 


tion, decarboxylation occurs to give acetaldehyde-ThPP. 
This compound reacts with oxidized lipoic acid to give acetyl 
dihydrolipoic acid, which in turn reacts with CoA to give 
dihydrolipoic acid and acetyl CoA (53-56). 

Wilson and Calvin (19), following their observation of 
glycolate accumulation at low CO2 pressure, suggested that 
the glycolyl moiety transferred by transketolase is the source 
of glycolic acid. We should now like to suggest specifically 
that the glycolaldehyde-ThPP compound formed in the first 
step of the transketolase or phosphoketolase reactions may 
undergo oxidation to give glycolyl CoA and, eventually, 
glycolate. This oxidation need not follow a pathway exactly 
analogous to the oxidation of acetaldehyde-ThPP, but we 
have shown it so in Figure 7. 

As mentioned earlier, during photosynthesis glycolate 
can be converted to the alpha and beta carbon atoms of PGA 
via carbon atoms 1 and 2 of the pentose in the carbon reduc- 
tion cycle. Thus it appears that the pathway from pentose 
phosphate to glycolate and glyceraldehyde phosphate should 
be reversible. The incorporation of glycolate via such a path- 
way would require an energy input, probably in the form of 
an activation by ATP. Finally, some state in the incorpora- 
tion pathway should involve equilibration with a symmetric 
intermediate because administration of glycolate-2-C^* to 
photosynthesizing plants leads to PGA labeled equally in 
the alpha and beta carbon atoms. We have indicated one 
such symmetric compoimd and there may be other possi- 

The formation of glycolyl CoA and reduced lipoic acid 
as shown in Figure 7 are hypothetical. If glycolyl CoA were 
formed, it seems likely that it would be an important inter- 
mediate in paths as yet unknown. In any event, if there is 
any conversion of carbon atoms number I and 2 of ketose 
to glycolic acid during photosynthesis, then an oxidation of 
the glycolyl fragment is required so that some cofactor, al- 
though not necessarily lipoic acid, must be reduced. 

Let us now attempt to explain the observation that 


labeled glycolate accumulates during photosynthesis with 
C^*02 when the CO2 pressure is reduced. 

1. Enzyme systems present in chloroplasts can bring 
about the oxidation of glycolate to glyoxylate with oxygen 
and the reduction of glyoxylate to glycolate with DPNH 
(57). If some steady-state relation between these two acids 
exists, it might well be shifted toward more glycolate at low 
CO2 pressures by the increase in the ratio of DPNH/DPN + 
that would result from the decreased utilization of TPNH 
for the carbon reduction cycle. Moreover, the oxidation of 
glycolate by O2 must in fact be limited in rate during photo- 
synthesis, or glycolate would not be seen at all. Possibly gly- 
colate is more effectively oxidized by some intermediate hy- 
droxy 1 or peroxide involved in the liberation of oxygen 
following the splitting of water during the primary act in 
photosynthesis. If so, such an intermediate oxidant may de- 
crease in concentration at low CO2 pressure because of re- 
combination with primary reductant that would build up, 
again as a result of decreased utilization by the carbon re- 
duction cycle. A decrease in the oxidant concentration would 
reduce the oxidation of glycolate. 

2. Low CO2 pressure might result in higher pH inside 
the chloroplasts. The phosphoketolase reaction, leading to 
acetyl CoA and involving the removal of OH" from gly- 
colaldehyde-ThPP, might be blocked, and the oxidation of 
the glycolaldehyde-ThPP to glycolyl CoA might be favored. 

3. If glycolyl CoA is formed and is a biosynthetic inter- 
mediate, the reactions in which it is used might require CO2 
analogous to the conversion of acetyl CoA to malonyl CoA 
in fatty acid biosynthesis. Low CO2 pressure could thus lead 
to an increased concentration of glycolyl CoA and permit its 
more rapid hydrolysis to glycolate. 

Tanner and co-workers (58,59) have recently proposed 
a direct route from CO2 to glycolic acid during photosyn- 
thesis. According to his scheme, CO2 is reduced by TPNH 
and MnCl~ to the radical CHO-. Two of these CHO- radi- 
cals are then condensed to give glyoxal, thence glycolic acid. 


This glycolic acid is then oxidized by 2 molecules of MnCl- 
(OH)2 (produced in the first step) to give glyoxylic acid. 
According to Tanner, the greater labeling of glycolic acid 
at low CO2 pressure during photosynthesis with C^^02 is 
due to the first step being first order with respect to the utili- 
zation of CO2 and the production of trivalent manganese, 
whereas the second step is second order with respect to the 
utilization of trivalent manganese. 

Whether or not Tanner's suggested route from CO2 to 
glycolic acid will be borne out by experiment remains to 
be seen. In all our experiments with C^*02, labeled glycolic 
acid has been a relatively minor product of the photosyn- 
thesis, except in those cases where the CO2 pressure has been 
permitted to drop to a very low level. Glycolic acid is some- 
what volatile, but it is a curious characteristic of this com- 
pound on paper chromatograms that, although 20 to 85 per 
cent may evaporate from the paper during development of the 
chromatogram, the remainder disappears only very slowly 
from the papers. This statement is based on measurement 
of radioactivity following chromatography of synthetic C^*- 
labeled glycolic acid. Thus it would seem that if a pathway 
leading directly from CO2 to glycolic acid (that is, with no 
isolable intermediates) were quantitatively important, we 
should have seen much more labeled glycolic acid following 
short periods of photosynthesis with C^^02, It could be that, 
under normal conditions of photosynthesis (say with 1 per 
cent CO2 in air), the reservoir size or concentration of gly- 
colic acid is very small, so that it would not appear to be 
strongly labeled, even though carbon from C^^02 enters it 
very rapidly. 

However, Moses and Calvin (44) conducted parallel 
experiments (3 minutes photosynthesis by Chlorella in the 
presence of C^*02 in one case and T2O in the other). The 
tritium-labeled glycolic acid accounted for more than 50 
per cent of the darkening of the radioautograph in the sub- 
sequent analysis by chromatography, whereas in the parallel 
experiment the glycolic acid contained less than 5 per cent 


of all the C^* found in compounds on the chromatograph. 
Thus the incorporation of hydrogen into nonexchangeable 
positions on glycolic acid seems to occur at ten times or 
more the rate of incorporation of C^* into the same com- 
pound. The simplest interpretation is that glycolic acid 
plays a much more important role in the transport of hy- 
drogen or reducing power than it does as an intermediate in 
carbon-compound formation from CO2. If any carbon di- 
oxide is reduced directly to glycolic acid during photosyn- 
thesis by Chlorella, it would seem to be a minor part of the 

A special role for glycolic acid in hydrogen transport is 
suggested by a combination of experimental findings from 
several laboratories. To Moses' finding of extremely rapid 
tritium labeling of glycolic acid and Tanner's implication 
of the role of glycolic acid with the requirement for man- 
ganese, we may add Delavin and Benson's report (60) of the 
light stimulation of the oxidation of glycolic acid with O2 
to glyoxylate and peroxide in isolated chloroplasts. Further, 
we must mention that manganese is thought by Kessler (61) 
to play some part in the formation of peroxide or O2 from 
water during the early stages of photosynthesis. Some form 
of peroxide is commonly postulated as an intermediate be- 
tween water and O2 during photosynthesis, and it may be 
that the plant has some mechanism for conserving the chem- 
ical potential energy that would be lost if peroxide were 
permitted to decompose to water and oxygen by a catalase 

The decrease in labeled glycolate in algae grown in 
Mn++ -deficient media (58,59) may be due to (1) some in- 
crease in the level of an intermediate in the oxygen-evolution 
pathway which is also capable of oxidizing glycolate to gly- 
oxylate (presumably Mn++ might be required for the break- 
down of this oxidant to O2); (2) a decrease in reduced pyri- 
dine nucleotide concentration, owing to impairment of the 
oxygen-evolving pathway; or (3) some enzymic requirement 


for Mn + + in the biosynthetic pathway from glycolaldehyde- 
ThPP to glycolate. 

Points (1) and (2) are related to the mechanisms sug- 
gested earlier for the effect of low CO2 pressure on glycolate 


As shown in Figure 7, acetyl phosphate can be formed 
from the carbon reduction cycle via the phosphoketolase 
pathway. This involves dehydration of the ThPP-acetalde- 
hyde compound derived from carbon atoms 1 and 2 of ketose 
phosphates. This route is especially attractive as a photo- 
synthetic pathway, since it conserves chemical energy and re- 
quires no oxidation or decarboxylation. Known enzyme sys- 
tems would readily convert the acetyl phosphate to acetyl 
CoA for fatty acid photosynthesis. 

Another pathway from the carbon reduction cycle to 
acetyl CoA could be via oxidative decarboxylation of pyruvic 
acid. This reaction is of the type we have earlier viewed as 
unlikely in photosynthesizing chloroplasts on grounds of 
economy. However, this economy takes on a different aspect 
if one considers the rapid formation of alanine, which we 
believe might be a reductive amination of phosphoenolpyru- 
vic acid derived from the carbon cycle (30). Our experiments 
indicate that about one-third of all NH4+ uptake occurs 
via this route. The resulting alanine must be used to a 
considerable extent in transamination reactions, resultine in 
the production of pyruvic acid. Although pyruvic acid is not 
labeled soon enough after the introduction of C^^02 to photo- 
synthesizing plants to permit us to consider it a precursor to 
alanine, it does become slowly labeled at later times. Thus 
pyruvic acid could be a product of transamination from ala- 
nine. The slow labeling of pyruvate may be because alanine 
has a very large reservoir, which does not saturate with C^* 
for some minutes. Once formed, the pyruvic acid cannot 


easily be converted back to PEPA. Rather, it must either go 
to malic acid via reductive carboxylation or be oxidized to 
acetyl CoA and CO2. 

The light-dark transient effect in C^^02 uptake during 
photosynthesis has often been observed (16,20). When the 
light is turned off, following a period of photosynthesis of 
algae with C^^02, labeled glutamic acid and citric acid ac- 
cumulate. One explanation of this effect has been given, 
based on the proposed formation of acetyl CoA by pyruvic 
acid oxidation. Lipoic acid in its oxidized form is required 
to accept the electrons in this oxidation. It was suggested 
that while the light is on this cofactor is kept mostly in its 
reduced state, dihydrolipoic acid. The reduced cofactor could 
not promote pyruvic acid oxidation. When the light is turned 
off and reducing power is no longer generated, the oxidized 
form of lipoic acid would be made, and the oxidation lead- 
ing to acetyl CoA would occur. Subsequent reactions, via the 
glyoxylate cycle, would then produce citric and glutamic 

However, if acetyl phosphate is formed by phosphoke- 
tolase during photosynthesis, a different explanation can be 
made. If we suppose that acetyl phosphate is still formed via 
phosphoketolase just after turning off the light, it will tend to 
accumulate. No reducing power or ATP is available for syn- 
thesis of fatty acids in the dark inside the chloroplasts. There- 
fore, acetyl phosphate will break down to free acetate, which 
will diffuse out of the chloroplast into the cytoplasm. There 
it will be used, via the glyoxylate cycle, in the synthesis of 
glutamic acid (62). 



^^^ Carbohydrates 

M onosaccharides 

The carbon reduction cycle (Figure 2) includes as in- 
termediate compounds the following sugar phosphates: 3- 
phosphoglyceraldehyde, dihydroxyacetone phosphate, fruc- 
tose- 1,6-diphosphate, fructose-6-phosphate, erythrose-4-phos- 
phate, sedoheptulose-l,7-diphosphate, sedoheptulose-7-phos- 
phate, xylulose-5-phosphate, ribulose-5-phosphate, ribose-5- 
phosphate, and ribulose-l,5-diphosphate. Besides these com- 
pounds, glucose phosphates are found to be very rapidly 
labeled in all plants in which we have studied the photosyn- 
thesis of carbon compounds from C^^02. When characterized, 
both glucose-6-phosphate and glucose- 1 -phosphate have been 
found. Other sugars found to be labeled somewhat more 
slowly in these experiments and identified as the free sugars 
following hydrolysis of the sugar monophosphate area include 
mannose and galactose. 

In virtually all the studies of the labeled products of the 
photosynthesis of carbon compounds from C^'*02 there has 
been found a striking absence of unphosphorylated mono- 
saccharides (14). This is hardly surprising, since photosyn- 


thesizing chloroplasts form phosphorylated sugars as inter- 
mediates in the carbon reduction cycle, since there is an abun- 
dance of ATP in the chloroplasts and since most known 
transformations of monosaccharides require phosphorylated 
forms of the sugars. Transformation of the phosphorylated 
sugars to the free sugars would for the most part result in 
a waste of chemical energy, for the sugar would then usually 
have to be phosphorylated again in reactions requiring ATP 
or UTP. Only when it becomes necessary to form a mole- 
cule that can be transported through the chloroplast mem- 
brane is it likely that a free sugar of relatively small molec- 
ular weight such as sucrose would be produced. 

A listing of various enzyme systems that appear to be 
responsible for the carbon reduction cycle has been delayed 
until now, since many of these biochemical steps are of in- 
terest in a discussion of carbohydrate synthesis. In Table 2 
there are listed the enzymes reported in the literature which 
appear to be responsible for steps of the carbon reduction 
cycle (Figure 2). Table 3 lists other enzymes which could 
account for subsequent steps in the synthesis of carbohydrates 
found to be labeled following relatively short periods of 
photosynthesis of algae with C^*02. 

We wish to emphasize that the finding of an enzyme in 
plant tissue does not, of course, prove that that particular re- 
action goes on in the photosynthesizing chloroplast either at 
all or in precisely the same way that it has been found to 
occur in vitro. Moreover, we would not consider the isolation 
of an enzyme with high catalytic activity a necessary condi- 
tion for believing that a given reaction may occur in vivo. 
The organization of the intact chloroplast inside the living 
cell and replete with all necessary natural cofactors and en- 
zymes is such that some steps which occur in vivo may prove 
extremely difficult to demonstrate in cell-free systems. None- 
theless, the isolation of a cell-free system, capable of carrying 
out a reaction that has been suspected on the basis of in vivo 
studies, is important corroborative evidence. 

The various enzymes listed in Tables 2 and 3, if present 
























































a o 

o -^ 

3 ' W 

rO X! so 'TT 


be I 





















r~- in in 
in ^ o C;^ CL- 

3 ^ 

< < 

§ I 
c 2 


r-- o^ r~- 1^ r~~ 

5 2 u 

3 O nj 

V3 ffi oi; 




















































"> "3 


CL wj CO to 









J= x: J3 

o o o 
c^ c^ c^ 

C fl C 

'a 'a 'a 'S, k?^ 

C/3 C/3 C/D C/2 J^i 



















' Oh 


Oh h 




5 + 
+ 3 


< ^ 


O Q 

"^r A 


CO ,-H 







Oh cC 

X X 



+ + + + 


Oh CLh 



-p r- 






fe C/5 








T T 













K ffi 

o o 








Oh Oh 
Q Q 

+ + 

Oh Oh 




fe V2 




rsi en 






TT ■*-* 

u' '^ 

t) -G 

U Q. 

-r o 

o -a 

o .2 

Oh H 








■^ en 

flj CT3 

h < 


to 5P 
O 2 

x: -^ 

























-3 a Q. ^ :5 







3 ^ 6 &■ 

oi ^ - £ I 

II II o 

CU Oh "S 

o in <" 

^ 0< II 

+-' f« t^ 

^ X ^ 

S o B 

X ^ ? 

0" . a 

vi ~ -^ 

^ Ih Qh 

2 be _" 

o 2 '^ 

3 j3 u 


a a 

II j: 


< "0 


•-0 " 

*H tin II 

^ Q 

g - Q 
° ..c« 

a ii •- 


«-> _ 

P O 



X cfl 

■5 ° 

s "^ 

be CO 

u o 


X •" 

*-^ CO 





o u 
t3 -2 








(^ *j 

X 2 

a -5 

CO D- 


X O 

o, X 

•- a. 

T3 ■ 

1 in 
un 1 

- lU 

^H CO 

CO 3 

Oh II 
►in m 


CO <L> 

.5 t« 

> 2 

< u 

II Oh 


Oh 3 

Q X 

3 .- 

.„ n 

V -C 

•^ CO 

X O 

O I 

X t^ 

a i 

in S 

















Oh .2 


" s 

«« 2 

U G 
1^ ® 

























CM — zl- --^ 
I- Ci >^ ho 00 

S "! S ^ , 

rt ^2 !/) .5 13 

5 -9 

S T3 





be (Tj 

3 Ji 

K& S^ 



















T + 











V Oh 



Oh T c^ 
O Oh g 

T S + 

Oh + O 

Q ^ ?:; 

o - 

o ^ 

rO Td- 


o ::; 

.22 3 

u S 


lo \0 


^ o 

O J3 

CL ' 
- \0 




a a 


o o 

CU Oh 






O O 

Oh Dh 

Q Q 

00 00 











































T Qi 


^ Q 







I~- 00 


>N CO 







2 '^ 

fli c 

Q bo 


-^ X 

>- « 


Ok 3 


CO >—' 






i) ^ 

bjD w 

II ^ 

so ^ 


■J in 
IS ^ 

X ^ 

CO " 

& Oh 
T) Q 




o id 

ho "eS 


„ X 

Oh a, 

Q S 

O x: 


















in chloroplasts, could account for virtually all the monosac- 
charide phosphates found to be significantly labeled with C^"* 
following a period of photosynthesis with C^*02 for several 
minutes in algae. Presumably there is present also another 
phosphohexose isomerase which catalyzes the conversion of 
fructose-6-phosphate to mannose-6-phosphate. 

Among the enzyme systems listed in Table 3 are several 
that utilize sugar nucleotides in the biosynthetic conversion 
of sugars. Such systems have been widely studied and have 
been discussed and reviewed elsewhere (88-90). Hassid and 
co-workers have widely studied the interconversions of sugars 
by these systems in higher plants and have summarized the 
interrelations of many of these systems in plants (91). Certain 
of these systems, which appear in Table 3, are particularly 
active in the early labeling of sugars in plants photosynthe- 
sizing with C^*02 and must be mentioned here, if only briefly. 

Buchanan et al. (15) identified uridine diphosphate glu- 
cose (UDPG) and uridine diphosphate galactose (UDPGal) 
in algae and found that the hexose moieties of these com- 
pounds were labeled with C^^ during short periods of 0^^*02 
photosynthesis even before sucrose. Thus the galactose found 
to be labeled in some experiments may be formed by the 
UDPG-UDPGal system. 

Disaccharides and polysaccharides 

As already indicated, when Chlorella pyrenoidosa photo- 
synthesizes in the presence of C^*02, sucrose is the first free 
sugar to be labeled to any extent. Benson (92) found that the 
radiocarbon in the fructose moiety of the sucrose, following 
photosynthesis of C'^02 by Chlorella, Scenedesmus, and soy- 
bean leaves, was greater than the radioactivity in the glucose 
moiety. Moreover, the difference between fructose and glu- 
cose became greater as the time of photosynthesis was de- 
creased. 1 he prior labeling of the fructose indicated that the 
glucose phosphate used in the synthesis of sucrose is formed 
from fructose phosphate. 


A study of the phosphorylated products of sfiort-term 
photosynthesis in C^^02 led to the discovery of a sucrose 
phosphate (93). The "hexose monophosphates" produced 
during photosynthesis in C^'^02 were treated with an inver- 
tase-free phosphatase preparation and subjected to paper 
chromatography. Although in most cases there were only 
minute traces of sucrose formed by this treatment, in sugar 
beet (5 minutes in Ci''02) there was an appreciable quantity. 
It was identified by cochromatography and enzymic hydroly- 
sis to glucose and fructose. 

When this "hexose monophosphate" sample was sub-^ 
jected to chromatography in ^butanol: picric acid: water, ra- 
dioactive areas corresponding to glucose-6-phosphate, fruc- 
tose-6-phosphate, sedoheptulose and mannose phosphates, 
and sucrose phosphate were obtained. The sucrose phosphate 
gave sucrose on phosphatase treatment, and on acid hydrolysis 
glucose and fructose phosphate were produced. The latter 
did not cochromatograph with fructose-6-phosphate. 

It appeared that in sucrose synthesis in green plants 
there are two possible mechanisms. Glucose- 1 -phosphate 
might react with fructose- 1 -phosphate to give sucrose phos- 
phate, which would be dephosphorylated to sucrose. Alter- 
natively, sucrose phosphate synthesis might be envisaged to 
occur through uridine diphosphate glucose (15), which be- 
comes labeled shortly before sucrose in kinetic experiments 
with €^^^02 (18). The uridine diphosphate glucose may be 
formed from glucose- 1 -phosphate by a UDPG pyrophosphory- 
lase (reaction 15, Table 3). This pathway is shown in Figure 
8 along with other pathways that may very likely occur dur- 
ing photosynthesis of carbohydrates from CO2. 

Leloir and Cardini (85) have isolated from wheat germ 
what appears to be two systems, one that catalyzes the reac- 
tion of fructose plus UDPG to give sucrose plus UDP, and 
a second that catalyzes the reaction UDPG plus fructose-6- 
phosphate to give sucrose phosphate plus UDP. Burma and 
Mortimer (94) have reported that with excised sugar beet 
leaves and leaf homogenates radioactive UDPG and sucrose 



CO2 Reduction- 


Sucrose P 





r:UDPG=::iUDPGal^;=rGal IP 



Figure 8. Biosynthetic pathways for photosyn- 
thesis of carbohydrates. 

were formed when radioactive glucose- 1 -phosphate, fructose- 
6-phosphate, and UTP were added. They propose a mecha- 
nism identical to that postulated by Buchanan except for the 
choice of fructose-6-phosphate as the precursor instead of 
fructose- 1 -phosphate. 

Not much is known about the formation of other poly- 
saccharides. There is a rapid labeling of unidentified polysac- 
charides during photosynthesis with C^^Oo. On the usual two- 
dimensional chromatogram, developed as described earlier, 
these compounds form what appears to be a homologous series 
of polyglucoses extending from the origin nearly to sucrose. 
The compound of this series closest to sucrose has been hy- 
drolyzed and found to contain only glucose. 



During photosynthesis by unicellular algae, it is not 
uncommon for as much as 30 per cent of the carbon dioxide 
taken up to be incorporated into fats. In Scenedesmus, for 
example, after 5 minutes in light in the presence of C'*- 
labeled carbon dioxide, 30 per cent of the fixed radioactivity 
is found in lipid materials. This incorporation of radiocarbon 
is greatly in excess of the rate of any synthesis that could 
take place in the dark and is an indication of the stimulation 
of fat production in the light. Fat synthesis requires a greater 
number of equivalents of reducing agents than does synthesis 
of carbohydrate or protein. Moreover, the composition of 
the chloroplasts includes a very high proportion of fat ma- 
terial. Since there is an abundance of reduced cofactors and 
ATP in the chloroplast, and since the end product, fat, is 
needed in the chloroplast, it is likely that much fat synthesis 
takes place in the chloroplast and is therefore to be consid- 
ered photosynthetic. 

Fatty acids 

All the well-known biosynthetic pathways to fatty acids 
require as a starting material acetate or acetyl CoA. We have 


already suggested under "Carboxylic Acids" four ways in 
which acetate, or acetyl CoA, could be made. These were: (1) 
splitting of malic acid to glyoxylate and acetate, (2) reduction 
of glycolic acid to acetate, (3) oxidation of pyruvic acid to 
acetyl CoA, and (4) dehydration and phosphoroclastic split- 
ting the postulated glycolyl-enzyme complex from transketo- 
lase reaction of the carbon reduction cycle to give acetyl phos- 
phate. We favor the last way as being the most likely. How- 
ever, if only the first three of these pathways are available, 
the third is probably the most important. 

However the acetate is formed, it is rapidly converted 
to fats in the light in algae. Experiments with Scenedesmus 
photosynthesizing in the presence of acetate- 1-C^^ and C^^02 
(14) demonstrated a light-accelerated incorporation of ace- 
tate into fats. A similar light-enhanced incorporation of ace- 
tate-2-C^* into lipids by Euglena was found by Lynch and 
Calvin (95). Sissakian (96) demonstrated the synthesis of 
higher fatty acids from labeled acetate in chloroplasts from 
sunflower plants. The utilization of free acetate in the light 
by chloroplasts is to be expected, since there is an abundance 
of ATP in the photosynthesizing chloroplasts for the conver- 
sion of acetate to acetyl phosphate and thence to acetyl CoA. 

The scheme of fatty acid synthesis proposed by Wakil 
and Ganguly (97) for the formation of fatty acids from 
acetyl CoA in animal tissues has been widely accepted. A 
similar pathway may exist in photosynthetic tissues. This 
pathway is incorporated in the hypothetical scheme of fat 
photosynthesis shown in Figure 9. Wakil (98) and Wakil and 
Ganguly (99) report that the first step in the synthesis from 
acetyl CoA is a carboxylation to give malonyl CoA. This step 
requires biotin and ATP, as well as Mn+ + . Malonyl CoA and 
acetyl CoA then condense to give acetoacetyl CoA, which 
then undergoes a series of reductive steps to give eventually 
butyryl CoA and carbon dioxide (97). 

Although the work of Ganguly and Wakil has been with 
animal tissues, it appears from the studies of Stumpf and 
co-workers (100-103) that similar systems of fatty acid syn- 


















r— "1 






o o 

^^ / 











































































thesis exist in plant tissues. The early stages of fat synthesis 
may well be similar in photosynthesizing chloroplasts to those 
known for other plant tissue and animals. The later stages 
and the fat products formed during photosynthesis in chloro- 
plasts are very likely different, since the chloroplast in all 
likelihood requires specialized fats for its operation. Benson 
and co-workers have identified a number of interesting com- 
pounds of glycerol phosphate and fatty acids as products of 
fat formation in green tissues. According to these workers, 
phosphatidyl glycerol is a major component of plant phos- 
pholipids. Moreover, they state that active transphosphatidyl 
action is observed during photosynthesis (104-106). 

Glycerol phosphate 

Alpha-D-glyceryl-1 -phosphate is presumably formed in 
chloroplasts during photosynthesis by direct reduction with 
TPNH of dihydroxyacetone phosphate. This compound 
could then be further converted to the polyglycerol phos- 
phates reported by Benson. The various glycerol phosphates 
would then presumably react with fatty acetyl CoA to pro- 
duce fats. Some of these postulated biosynthetic routes are 
shown in Figure 9. 



"^J^ Pigments 

Of major importance among the biosynthetic pathways 
of the chloroplast must be those leading to photosynthetic 
pigments. Akhough some of these may vary from one organ- 
ism to another, all organisms must be capable of making at 
least one of the chlorophylls, carotenoids, and hematin pig- 
ments. During photosynthesis the simple precursor molecules 
for these synthetic paths are available from the carbon reduc- 
tion cycle, whereas the reduced pyridine nucleotides and 
ATP are of course at high levels in the chloroplast. 

Carotenoids and phytol 

The starting point for the synthesis of carotenoids and 
phytol, as well as steroids and terpenes, is acetyl CoA. In the 
previous sections we discussed routes from the carbon reduc- 
tion cycle to acetyl CoA. These are shown in Figures 7 and 9. 

The biosynthetic paths to terpene compounds have been 
much clarified in recent years by work from the laboratories 
of Lynen (107), Bloch (108), Folkers (109), and Popjak (110). 
Successive condensations of acetyl CoA give acetoacetyl CoA 
and then y8-hydroxy-^-methyl-glutaryl (or crotonyl)-CoA 


(HMG-CoA). The HMG-CoA is then reduced to give meva- 
lonic acid (Figure 9). Further steps along the biosynthetic 
path are shown in Figure 10. Pyrophosphorylation and de- 
carboxylation of mevalonate give isopentenyl-pyrophosphate, 
the biological isoprene unit. 

According to Lynen, isopentenyl-pyrophosphate units 
then condense to give, successively, Cio, C15, and C20 com- 
pounds, as shown in Figure 10. Hydrogenation of the C20 
compound could presumably lead to phytol, an alcohol that 
forms the phytyl tail of chlorophyll. Dimerization of the C15 
compound, farnesyl pyrophosphate, gives squalene, the pre- 
cursor for steroids. We might expect the C20 compound, 
geranylgeranyl pyrophosphate, to undergo a similar conden- 
sation to give C40 compounds, which could in turn be con- 














Figure 10. The biosynthesis of carotenoids. 

(* For details see Figures 7 and 9.) 


verted to carotenoids. Stanier (HI) has reported evidence 
indicating that the initial compound in this series is phytoene 
or tetrahydrophytoene (see Figure 10). 

Present evidence indicates that conversion of the C40 
compound formed from the condensation, to carotenoids, 
involves a number of dehydrogenations, and finally ring 
closure at the ends of the molecule. The various oxygen-con- 
taining carotenoid compounds are probably formed by oxida- 
tions, hydrations, etc. The structures of a great many of these 
compounds, both intermediates and end products, have been 
established in the laboratories of Karrer (112), Zechmeister 
(113), Inhoffen (114), Weedon (115), and others. 

Chlorophyll and heme 

The pathways to porphyrin compounds have been re- 
cently reviewed by Granick (116,117), Shemin (118), Rim- 
ington (119), and Bogorad (120). Some of the key steps from 
these paths are shown in Figure 11. Glycine and succinate 
formed from the carbon reduction cycle are the starting com- 
pounds for the syntheses of these pigments. Glycine may be 
formed from serine, which in turn is probably synthesized 
from 2-phosphoglycerate, formed from the 3-phosphoglycerate 
of the cycle (see the section on Amino Acids). Alternatively, 
glyoxylate may be transaminated to give glycine. The deriva- 
tion of this glyoxylate from the carbon reduction cycle is not 
known for certain, but is probably related to the formation 
of glycolic acid (see the section on Carboxylic Acids). Thus 
glycolate formed by oxidation of the glycolyl fragment from 
the sugar phosphate transketolase system could be further 
oxidized to glyoxylic acid. A hypothetical split of malate 
could lead to acetate and glyoxylate. 

If the chloroplast contained isocitritase, both succinate 
and glyoxylate could be formed by the same reaction on iso- 
citrate. The isocitrate would in this case come from acetyl 
CoA and oxalacetate condensation, via citrate. Oxalacetate 


























F. + + 





Vi = -CH=CH2 

Pr = -CH2-CH2-CO2H 

Mt' -CH3 



Figure 11. The biosynthesis of porphyrins. 

is formed from the cycle by carboxylation of phosphoenolpy- 
ruvate, derived from phosphoglycerate. 

Another, and perhaps more likely route to succinate is 
via reductive carboxylation to form malate, dehydration, and 
reduction of malate to give succinate. 


As shown in Figure 11, condensation of glycine with suc- 
cinic acid gives S-amino levulinic acid, which in turn con- 
denses with itself to make a substituted pyrrole ring (por- 
phobilinogen). Condensations and isomerizations, the exact 
mechanisms of which are not known, lead to the formation 
of the tetrapyrrole structure of uroporphyrinogen(III) from 
four porphobilinogen molecules. 

The conversion of uroporphyrinogen to protoporphyrin 
requires a number of decarboxylations of the substituent acyl 
groups, oxidation of two of these groups to vinyl groups, and 
dehydrogenation and aromatization of the pyrrole rings and 
the methylene bridges connecting them. 

Protoporphyrin-9 is an important branching point: in- 
corporation of Fe++ leads to heme and thence to the various 
hematin pigments, whereas incorporation of Mg++ ion leads 
ultimately to the synthesis of the chlorophylls. The latter 
pathway must first accomplish the formation of the fifth 
ring and the partial saturation of one of the pyrrole rings. 

Finally the phytol alcohol, probably formed as shown 
in Figure 10, is attached to the pigment molecule as a phytyl 
group, and the synthesis of chlorophyll is complete. At some 
time, before or after this step, the alterations needed to make 
the various forms of chlorophyll, and to incorporate it into 
the structure of the photosynthetic apparatus are completed. 


'^tf' Aromatic nuclei 

The shikimic acid pathway for the biosynthesis of aro- 
matic compounds, including amino acids, from carbohydrates 
has been well established by the work of Davis (38) and his 
collaborators, who used biochemical mutants of E. coli. 
Without going into the details of this pathway, we may point 
out that the starting materials are phosphoenolpyruvate, 
which is readily available as a photosynthetic intermediate, 
and D-erythrose-4-phosphate, which is also an intermediate of 
the carbon reduction cycle. Presumably, therefore, the syn- 
thesis of aromatic amino acids in photosynthesizing plants 
would follow a pathway similar to the shikimic acid path- 
way. The first step in that pathway is the condensation of 
phosphoenolpyruvate with erythrose-4-phosphate to give a 
seven-carbon compound which has been identified as 2-keto- 
3-deoxy-D-araboheptonic acid-7-phosphate. This intermediate 
subsequently undergoes ring closure to give dehydroquinic 
acid. Rearrangements via a number of additional steps gives, 
eventually, phenylalanine and tyrosine. Higuchi (121) has 
summarized some of the reasons for believing that the shi- 


kimic acid pathway does occur in higher plants. For example, 
shikimic acid is of widespread occurence, and some of the 
enzymes of the pathway have in fact been found in higher 
plants. Neish (122) has further reviewed evidence for the shi- 
kimic acid pathway in plants. 




^^^ products 

As we learn more about the capabilities of the chloro- 
plast to form compounds from carbon during photosynthesis, 
we come closer to the conclusion that the chloroplast, as it 
exists in the living, undisturbed cell, is a self-sufficient factory 
capable of producing essentially all the materials required 
for its replenishment. Thus it appears to be able to make 
all kinds of sugars, polysaccharides, protein, fats, pigments, 
enzymes, and cofactors. In addition to this, it produces for 
export to the cytoplasm reserves of organic compoimds. 
These are probably sugars, glycolic acid, and other neutral, 
relatively small, molecules which can be readily transported 
through the chloroplast membrane. Until more is known 
about the development and formation of chloroplasts, we can- 
not say just when it gains this complete synthetic ability. No 
doubt there are early stages in the development of chloro- 
plasts in which it must be built from cytoplasmic materials 
derived in turn from already-functioning chloroplasts. There 
is no reason to suppose the chloroplast functions without nu- 
clear control, even though it does not appear to have a nu- 
cleus of its own. Presumably it is possible for RNA mole- 
cules to move in and out of the chloroplast in some way. It 


cannot be said at the moment whether or not the chloroplast 
is capable of synthesizing nuclear material. It would seem 
likely, however, that the chloroplast can synthesize pu- 
rines and pyrimidines, coenzymes, and nucleotide materials 
needed for the continued functioning of the chloroplast as 
a self-sufficient biosynthetic factory. If, as we now think, 
protein synthesis and enzyme synthesis occur in the chloro- 
plast, then either the chloroplast must obtain a store of RNA 
molecules at its initial construction or else such molecules 
must be able to travel back and forth from the chloroplast 
to the cytoplasm. 

In conclusion, we should say that the point of view of 
the ability of the chloroplast to carry out photosynthetic 
formation of many compounds is a departure from the view 
held only a few years ago. It was then thought that the primary 
function of photosynthesis was to form carbohydrate only. 
This carbohydrate was then thought to be used by the cyto- 
plasm in the synthesis of all other compounds. Of course, the 
chloroplast must stipply the carbohydrate and reducing power 
for the cytoplasmic synthesis. It now appears that chloroplasts 
also synthesize a complete spectrum of biochemical products, 
all of which might reasonably be considered to be photosyn- 
thetic products. Finally, as we learn more about the photo- 
synthetic paths to these products, we are impressed not 
merely by their complexity but much more by the economy 
with which both energy and material are utilized. 


^^^ References 

1. Bassham, J. A., A. A. Benson, L. D. Kay, A. Z. Harris, A. T. 
Wilson, and M. Calvin, /. Am. Chem. Soc, 76, 1760 (1954). 

2. Bassham, J. A., and M. Calvin, The Path of Carbon in Pho- 
tosynthesis, Prentice-Hall, Inc., Englewood Cliffs, N.J., 1957. 

3. Bassham, J. A., and M. Kirk, Biochim. Biophys. Acta, 43, 
447 (1960). 

4. Norris, L. T., R. E. Norris, and M. Calvin, /. Exptl. Botany, 
16, 64 (1955). 

5. Moses, v., and M. Calvin, Proc. Natl. Acad. Sci. U.S., 44, 
260 (1958). 

6. Axelrod, B., and R. S. Bandurski, /. Biol. Chem., 204, 939 

7. Arnon, D. I., Science, 116, 635 (1952). 

8. Gibbs, M., Nature, 170, 164 (1952). 

9. Stumpf, P. K., /. Biol. Chem., 176, 233 (1948). 

10. Stumpf, P. K., /. Biol. Chem., 182, 261 (1950). 

11. Stepka, W., A. A. Benson, and M. Calvin, Science, 108, 304 

12. Benson, A. A., J. A., Bassham, M. Calvin, T. C. Goodale, 
V. A. Haas, and W. Stepka, /. Am. Chem. Soc, 72, 1710 

13. Bassham, J. A., A. A. Benson, and M. Calvin, /. Biol. Chem., 
185, 787 (1950). 


14. Calvin, M., J. A. Bassham, A. A. Benson, V. Lynch, C. 
Ouellet, L. Schou, W. Stepka, and N. E. Tolbert, Symposia 
Soc. Exptl. Biol, 5, 284 (1951). 

15. Buchanan, J. G., V. Lynch, A. A. Benson, D. Bradley, and 
M. Calvin, /. Biol. Chem., 203, 935 (1953). 

16. Calvin, M., and P. Massini, Experientia, 8, 445 (1952). 

17. Calvin, M., Proc. Third Intern. Biochem. Congress, Brussels, 
1955, p. 211. 

18. Benson, A. A., S. Kawaguchi, P. M. Hayes, and M. Calvin, 
J. Am. Chem. Soc, 74, 4477 (1952). 

19. Wilson, A. T., and M. Calvin, J. Am. Chem. Soc, 77, 5948 

20. Bassham, J. A., K. Shibata, K. Steenberg, J. Bourdon, and 
M. Calvin, /. Am. Chem. Soc, 78, 4120 (1956). 

21. Park, R. B., and N. G. Pon, /. Mol. Biol, 3, 1 (1961). 

22. Tolbert, N. E., in The Photochemical Apparatus: Its Struc- 
ture and Function, Brookhaven Symposia in Biology, Office 
of Technical Services, Department of Commerce, Washing- 
ton, D.C., 1958, vol. 11, p. 271. 

23. Aronoff, S., A. A. Benson, W. Z. Hassid, and M. Calvin, 
Science, 105, 664 (1947). 

24. Benson, A. A., M. Calvin, V. A. Haas, S. Aronoff, A. G. Hall, 
J. A. Bassham, and J. W. Weigl, in James Franck and W. E. 
Loomis (eds.). Photosynthesis in Plants, Iowa State College 
Press, Ames, 1949, pp. 381-401. 

25. Nichiporovich, A. A., "Tracer Atoms Used to Study the 
Products of Photosynthesis as Depending on the Conditions 
in Which the Process Takes Place," Proc. First Geneva Conf. 
on Peaceful Uses of Atomic Energy, 1955. 

26. Andreyeva, T. F., Doklady Akad. Nauk SSSR, 78, 1033 

27. Voskrenskaya, N. R., Doklady Akad. Nauk SSSR, 93, 911 

28. Nezgovorova, Fiziol. Rastenii Akad. Nauk SSSR, 6, 451 

29. Sissakian, N. M., Proc. Second Intern. Conf. on Peaceful 
Uses of Atomic Energy, Geneva, 1958, Part 2, vol. 25, p. 159. 

30. Smith, D. C, J. A. Bassham, and M. Kirk, Biochim. Biophys. 
^c<a, 48, 299 (1961). 


31. Moses, v., O. Holm-Hansen, J. A. Bassham, and M. Calvin, 
/. Mol. Biol., 1, 21 (1959). 

32. Van der Meulen, P. Y. F., and J. A. Bassham, /. Am. Chem. 
Soc, 81, 2233 (1959). 

33. Holm-Hansen, O., N. G. Pon, K. Nishida, V. Moses, and M. 
Calvin, Physiol. Plantarum, 12, 475 (1959). 

34. Warburg, O., Science, 128, 68 (1958). 

35. Barker, H. A., R. M. Wilson, and A. Munch-Petersen, Fed- 
eration Proc, 16, 151 (1957). 

36. Munch-Petersen, A., and H. A. Barker, /. Biol. Chem., 230, 
649 (1958). 

37. Barker, H. A., R. D. Smith, R. M. Wilson, and H. Weissbach, 
/. Biol. Chem., 234, 320 (1959). 

38. Davis, B. D., Arch. Biochem. Biophys., 78, 497 (1958). 

39. Dekker, E. E., Biochim. Biophys. Acta, 40, 174 (1960). 

40. Virtanen, A. I., and P. K. Hietala, Acta Chem. Scand., 9, 175 

41. Benson, A. A., and M. Calvin, /. Exptl. Botany, 1, 63 (1951). 

42. Schou, L., A. A. Benson, J. A. Bassham, and M. Calvin, 
Physiol. Plantarum, 3, 487 (1950). 

43. Tolbert, N. E., and L. P. Zill, /. Biol. Chem., 222, 895 (1956). 

44. Moses, V., and M. Calvin, Biochim. Biophys. Acta, 33, 297 

45. Horecker, B. L., P. Z. Smyrniotis, and H. Klenow, /. Biol. 
Chem., 205, 661 (1953). 

46. Racker, E., G. de la Haba, and I. G. Leder, Arch. Biochem. 
Biophys., 48, 238 (1954). 

47. Heath, E. C, J. Hurwitz, B. L. Horecker, and A. Ginsburg, 
/. Biol. Chem., 231, 1009 (1958). 

48. Schram, M., and E. Racker, Nature, 179, 1349 (1957). 

49. Breslow, R., /. Am. Chem. Soc, 80, 3179 (1958). 

50. Breslow, R., /. Cellular Comp. Physiol., 54, Suppl. 1, 100 

51. Krampitz, L. O., /. Cellular Comp. Physiol, 54, Suppl. 1, 
101 (1959). 

52. Krampitz, L. O., G. Greull, C. S. Miller, J. B. Bicking, H. R. 
Skeggs, and J. M. Sprague, /. Am. Chem. Soc, 80, 5893 

53. Reed, L. J., B. D. DeBusk, I. C. Gunsalus, and C. S. Horn- 
berger, Jr., Science, 114, 63 (1951). 


54. Reed, L. J., Adv. in Enzymol, 18, 319 (1956). 

55. Gunsalus, I. C, L. S. Barton, and W. Gruber, ]. Am. Chem. 
Soc, 78, 1763 (1956). 

56. Gunsalus, I. C, in W. D. McElroy and B. Glass (eds.), Mech- 
anisms of Enzyme Action, Johns Hopkins Press, Baltimore, 
1951, Vol. I, p. 366. 

57. Zelitch, I., and G. A. Barber, Plant Physiol., 35, 623 (1960). 

58. Tanner, H. A., T. E. Brown, C. Eyster, and R. W. Treharne, 
Ohio J. Sci., 60, 231 (1960). 

59. Tanner, H. A., T. E. Brown, C. Eyster, and R. W. Treharne, 
Biochem. Biophys. Research Comm., 3, 205 (1960). 

60. Delavin, L. A., and A. A. Benson, The Photochemical Appa- 
ratus: Its Structure and Function, Brookhaven Symposia in 
Biology, Office of Technical Services, Department of Com- 
merce, Washington, D.C., 1958, vol. 11, p. 259. 

61. Kessler, E., in H. Gaffron et al. (eds.). Research in Photo- 
synthesis, Interscience Publishers, Inc., New York, 1957, p. 

62. Kornberg, H. L., Proc. Fourth Intern. Biochem. Congr., 
Vienna, 1958, vol. 13, p. 251. 

63. Quayle, J. R., R. C. Fuller, A. A. Benson, and M. Calvin, 
/. Am. Chem. Soc, 76, 3610 (1954). 

64. Mayaudon, J., A. A. Benson, and M. Calvin, Biochim. 
Biophys. Acta, 23, 342 (1957). 

65. Weissbach, A., and B. L. Horecker, Federation Proc, 14, 
302 (1955). 

66. Weissbach, A., B. L. Horecker, and J. Hurwitz, /. Biol. 
Chem., 218, 795 (1956). 

67. Weissbach, A., P. Z. Smyrniotis, and B. L. Horecker, /. Am. 
Chem. Soc, 76, 3611 (1954). 

68. Jakoby, W. G., D. O. Brummond, and S. Ochoa, /. Biol. 
Chem., 21S, 811 (1956). 

69. Racker, E., Arch. Biochem. Biophys., 69, 300 (1957). 

70. Arnon, D. E., L. L. Rosenberg, and F. R. Whatley, Nature, 
173, 1132 (1954). 

71. Rosenberg, L. L., and D. I. Arnon, /. Biol. Chem., 217, 361 

72. Fuller, R. C, and M. Gibbs, Plant Physiol, 31, xxxi (1956). 

73. Tewfic, S., and P. K. Stumpf, /. Biol. Chem., 192, 519 (1951). 

74. Tewfic, S., and P. K. Stumpf, Am. J. Botany, 36, 567 (1949). 


75. Hough, L., and J. K. N. Jones, /. Chem. Soc, 1953, 342. 

76. Racker, E., Nature, 175, 249 (1955). 

77. Racker, E., and E. A. R. Schroeder, Arch. Biochem. Biophys., 
74, 326 (1958). 

78. Srere, P. A., J. R. Cooper, V. Klybas, and E. Racker, Arch. 
Biochem. Biophys., 59, 535 (1955). 

79. Hurwitz, J., A. Weissbach, B. L. Horecker, and P. Z. Smyrn- 
iotis, /. Biol. Chem., 218, 769 (1956). 

80. Weissbach, A., P. Z. Smyrniotis, and B. L. Horecker, /. Am. 
Chem. Soc, 76, 5572 (1954). 

81. Ramasarma, T., and K. V. Giri, Arch. Biochem. Biophys., 
62, 91 (1956). 

82. Morita, S., S. Makamura, and T. Ito, Ochanomizu Joshi 
Daigaku Shizenkagaku Hokoku, 4, 68 (1955). 

83. Sidbury, J. B., and V. A. Najjar, /. Biol. Chem., 227, 517 

84. Ginsberg, V., /. Biol. Chem., 232, 55 (1958). 

85. Leloir, L. F., and C. E. Cardini, /. Biol. Chem., 214, 157 

86. Neufeld, E. F., V. Ginsberg, E. W. Putman, D. Fanshier, and 
W. Z. Hassid, Arch. Biochem. Biophys., 69, 602 (1957). 

87. Kalckar, H. M., Biochim. Biophys. Acta, 12, 250 (1953). 

88. Leloir, L. F., Proc. Third Intern. Biochem. Congr., Brus- 
sels, 1955, p. 154. 

89. Baddiley, J., and J. G. Buchanan, Quart. Revs. (London), 
12, 152 (1958). 

90. Kalckar, H. M., Adv. in Enzymol, 20, 111 (1958). 

91. Hassid, W. Z., E. F. Neufeld, and D. S. Feingold, Proc. Natl. 
Acad. Sci. U.S., 45, 905 (1959). 

92. Benson, A. A., Arch. Biochem. Biophys., 32, 223 (1951). 

93. Buchanan, J. G., Arch. Biochem. Biophys., 44, 140 (1953). 

94. Burma, D. P., and D. C. Mortimer, Arch. Biochem. Biophys., 
62, 16 (1956). 

95. Lynch, V. H., and M. Calvin, Ann. N.Y. Acad. Sci., 56, 890 

96. Sissakian, N. M., and B. P. Smirov, Biokhimiya, 21, 275 

97. Wakil, S. J., and J. Ganguly, /. Am. Chem. Soc, 81, 2597 

98. Wakil, S. J., /. Am. Chem. Soc, 80, 6465 (1958). 


99. Wakil, S. J., and J. Ganguly, Federation Proc, 18, 346 (1959). 

100. Stumpf, P. K., Federation Proc, 18, 329 (1959). 

101. Hatch, M. D., and P. K. Stumpf, Pacific Slope Biochemical 
Conference Abstract, Paper 28, 1960. 

102. Mudd, J. B., and P. K. Stumpf, Pacific Slope Biochemical 
Conference Abstract, Paper 29, 1960. 

103. Barron, E. J., and P. K. Stumpf, Pacific Slope Biochemical 
Conference Abstract, Paper 30, 1960. 

104. Benson, A. A., and B. Maruo, Biochim. Biophys. Acta, 27, 
189 (1958). 

105. Benson, A. A., J. F. G. Wintermans, and R. Wiser, Plant 
Physiol., 34, 315 (1959). 

106. Benson, A. A., and E. H. Strickland, Biochim. Biophys. Acta, 
41, 328 (1960). 

107. Lynen, F., B. W. Agranoff, H. Eggerer, U. Henning, and 
E. M. Moslein, Angew. Chem., 71, 657 (1959); Eggerer, H., 
and F. Lynen, Ann. Chem. Liebigs, 630, 58 (1960). 

108. Rilling, H. C, and K. Block, /. Biol. Chem., 234, 1424 (1959). 

109. Folkers, K., C. H. Shunk, B. O. Linn, F. M. Robinson, P. E. 
Wittreich, J. W. Huff, J. L. Gilfillan, and H. R. Skeggs, 
Ciba Foundation Symposium on Biosynthesis of Terpenes 
and Sterols, J. A. Churchill, Ltd., London, 1959, Sec. 20. 

110. Popjak, G., Ciba Foundation Symposium on Biosynthesis of 
Terpenes and Sterols, J. A. Churchill, Ltd., London, 1959, 
Sec. 148. 

111. Stanier, R. Y., Harvey Lectures, 54, 219 (1960). 

112. Entschel, R., and P. Karrer, Helv. Chim. Acta, 42, 466 (1959). 

113. Zechmeister, L., Fortschr. Chem. org. Naturstoffe, 15, 31 
(1958); 18, 223 (1960). 

114. Inhoffen, H. H., and D. Erdmann, Ann. Chem. Liebigs, 598, 
51 (1956). 

115. Akhtar, M., and B. C. L. Weedon, /. Chem. Soc. (London), 
1959, 4058; Davis, J. B., and B. C. L. Weedon, Proc. Chem. 
Soc, 1960, 182. 

116. Granick, S., Proc. Fifth Intern. Biochem. Congr., Moscow, 
August 1961, Symposium VI, in press. 

117. Granick, S., and D. Mauzerall, in D. M. Greenberg (ed.), 
Chemical Pathways of Metabolism, Academic Press, Inc., 
New York, 1961, vol. 2. 

118. Shemin, D., Ergeb. Physiol. 49, 299 (1957). 


119. Rimington, C, Ann. Rev. Biochem., 26, 561 (1957). 

120. Bogorad, L., in M. B. Allen (ed.), Comparatwe Biochemistry 
of Photoreactive Systems, Academic Press, Inc., New York, 
1960, p. 227. 

121. Higuchi, T., Proc. Fourth Intern. Biochem. Congr., Vienna, 
1958, vol. 2, p. 161. 

122. Neish, A. C, Ann. Rev. Plant Physiol, 11, 55 (1960). 



EXPERIENTIA VOL. Vni/12. 1952 - p. 445 


The Path of Carbon in Photosynthesis' 

(XX. The Steady State) 
By M. Calvin and P. Massini^ Berkeley, Cal. 

Photosynthesis, the process by which green plants 
are able to capture electromagnetic energy in the form 
of sunlight and transform it into stored chemical energy 
in the form of a wide variety of reduced (relative to 
carbon dioxide) carbon compounds provides the only 
major soufce of energy for the maintenance and pro- 
pagation of all life. For this and other reasons, the 
study of the nature of this process has been a very 
attractive area for many years and a wide variety of 
scientific interest and backgrounds have been brought 
to bear upon it. These range from the purely biological 
to the strictly physical with the biochemical and phy- 
sicochemical area lying between. Important contri- 
butions to the understanding of the phenomenon have 
come from all these areas, but in spite of the enormous 
amount of work and study that has gone into the prob- 
lem, relatively little is known, or rather understood, 
about the fundamental character of the process even 
today. It is perhaps pardonable that one engaged in 
studies in this area tends to the conclusion that most of 
the knowledge has been acquired in the relatively recent 
past. Discounting that tendency, it still seems fair to 
say that we have only just begun in the last decade or 
so to gain some understanding of the intimate details 
by which the basic process represented in the overall 

CO, + H,0 


-> 0,+ {CH,0), 

- Energy 

has come to be understood. The recognition of this 
overall reaction as written, to represent the basic nature 
of the process of photosynthesis, and, further, that its 
reversal represents the basic reaction of respiration is, 
of course, an old one. 

As a result of more recent study, it has been possible 
to separate the process of photosynthesis into two dis- 
tinct and separate parts. The general features of this 

^ The work described in this paper was sponsored by the U.S. 
Atomic Energy Commission. 

■ Radiation Laboratory and Department of Chemistry, Univer* 
sity of California, Berkeley. Fellow of the Swiss Foundation, 
iStiftung fiir Stipendien auf dem Gebiete der Chemie», 1951-1952. 

separation may be represented in the following chart 
(Fig. 1). The essential feature of the separation is the 
independence of the photochemical part of photosyn- 
thesis from the carbon dioxide reduction part. We shall 
not here even try to outline all of the various forms of 
evidence which have been adduced in support of such 
a scheme but only to point out additional bits which 
have been added in recent years and particularly those 
which stem from our own work'. 




Fig. 1. 

The scheme itself is an outgrowth of proposals of 
some fifteen years ago by Van Niel* resulting from his 
studies of the comparative biochemistry of photosyn- 
thesis. More recently, the photochemical apparatus has 
been shown to be separable from the rest of the plant 
by the experiments of Hill'. 

He was able to make preparations of chloroplasts 
and chloroplastic fragments which, upon illumination 
in the presence of suitable oxidizing agents other than 
carbon dioxide, were able to evolve molecular oxygen. 
Still more recently, Ochoa an others* were able to 
demonstrate that these same preparations were capable 
of using coenzyme I and II (D.P.N, and T.P.N.) as 

' M. Calvin and A. A. Benson, Science 107, 476 (1948). - A. A. 
Benson and M. Calvin, Cold Spring Harbor Symp. quant. Biol. tS, 
6 (1948). - M. Calvin and A. A. Benson, Science lOS, 140 (1949). 

» C. 3. Van Niel, P*o/osy>i(A«si5 in P/anb, Chapter 22 (Iowa State 
College Press, Ames, Iowa, 1949), pp. 437-495. 

' R. Hill, Nature 139, 881 (1947); Proc. roy. Soc. (London) [BJ 
127, 192 (1939).- R. Hill and R. Scarisbrick, Nature H6. 61 (1940). 

• W. VisHNlAC and S. Ochoa, J. Biol. Chem. 19S, 75 (1952). - 
D. I. Arnon, Nature H7, 1008 (1951). - L. J. Tolmach, Arch, Bio- 
chem. Biophys. 33, 120 (1951). 



u viN and r. Massisi: TIip Tnth of C.nrl.nn in Photosynthesis 


suitable oxidizing agents leading to the evolution of 
oxygen. Furthermore, the experiments of Ruben* 
showed that the molecule of oxygen evolved in photo- 
synthesis had its proximate origin in the oxygen of 
the water molecule and that the oxygen atom associ- 
ated with the carbon dioxide must first pass through 
water before arrivingatgaseousoxygcn. From the chart 
it may be seen that the ultimate result, then, of- the 
photochemical reaction initiated by the absorption of 
light by the chlorophyll molecule is the division of the 
water molecule into an oxidized part which ultimately 
leads to molecular oxygen and some reduced parts 
represented in the chart by [H], 

This reduced part [H] we have called "reducing 
power" because as yet it is not possible to state specifi- 
cally what form or forms it may be in. This reducing 
power is capable of reducing carbon dioxide in the 
absence of light; that is to say, that the reduction of 
carbon dioxide itself is a dark reaction. This was indi- 
cated first in the earlier experiment of McAlister" in 
which he was able to show that following a period of 
photosynthesis a number of plants continued to absorb 
carbon dioxide for a short period (seconds to minutes) 
after cessation of illumination. We were able to demon- 
strate this in an even more direct and uneciuivocal 
fashion and generalize it for all plants so far tried when 
we were able to show that not only did all of these 
plants absorb (|uantities of carbon dioxide in the dark 
after illumination but that the products formed in the 
dark were (jualitatively and under certain conditions 
quantitatively similar to those formed in a fairly com- 
parable light period'. The method used for this demon- 
stration was the same as those to be described later in 
the review. The lifetime in the dark of this reducing 
power which is generated by light is also of the order 
of seconds to minutes and almost certainly corresponds 
to a concentration of one or more definite chemical 
species. It is quite conceivable, as mentioned earlier, 
that some of it might be in the form of reduced coen- 

Very recently it has been reported* that both the 
higher plants and isolated chloroplasts emit a chemi- 
luminiscence following cessation of illumination. This 
chemiluminiscence has a decay time which corresponds 
very closely to that which we have observed for the 
reducing power. In fact, it would seem almost surely 
to represent the reversal of the conversion of electro- 
magnetic into chemical energy, namely, the transfor- 
mation of at least some of the chemical energy stored 
in the reducing power into the electromagnetic energy 
of luminiscence. Furthermore, the luminiscence is re- 

' S. UruKN, M. Randaij., M. 11. K\men, .ind J. Hvor. J. A\\\. 
Chcin. Soc. C3, 877 (1941). 

' K. U. McAr.isTER and J. MvF.RS, J. Smithsonian Insl. I'uM 
(Misc. Coll.) c, aa (1940). 

' M. Calvin, J. Chcui. Uducation J6, 030 (lOl'J). 

* B. L. STREHi.tR and W. .\rnoi.u, J. Gen. Physiol. 34, sua 
(lull). - H. I.. STRtHUF.R, .^rch. BiochoiTi. Hiophys. 34, M9 (19:.l) 

duced by the presence of carbon dioxide in those cases 
in which the carbon dioxide fixing system is still pre- 
sent. However, when the carbon dioxide system has 
been removed, as is true in the case of chloroplasts, the 
luminiscence becomes independent of carbon dioxide. 
While it thus appears that the unique problem of 
photosynthesis lies in the right hand half of thechart 
of Figure 1, the present discussion will be limited to 
the other side of the chart, that is, the path through 
which carbon passes on its way from carbon dioxide to 
all the raw materials of the plant. It is essentially a 
study of what we now believe to be entirely dark 
reactions and might best be characterized as phyto- 
synthesis. This area not only has a great interest for its 
own sake but would almost certainly cast some light 
upon the nature of the reducing agents which arrive 
from the photochemical part of the reaction and drive 
the carbon cycle toward reduction. The reason for this 
particular interest lies in the fact that we have, in recent 
years, come into possession of a tool which is especially 
suited for this study, namely, labeled carbon atoms in 
the form of a radioactive isotope of carbon, O*. All of 
the results that will be described later were made 
possible through the use of this labeled carbon dioxide. 
With such a labeled molecule available, the design of 
an experiment for determining the sequence of com- 
pounds into which the carbon atoms of carbon dioxide 
may pass during the course of their incorporation in the 
plant is, in its first phase, a straightforward one. 



We may visualize the problem in terms of the chart 
in Figure 2 in which the green leaf is represented 
schematically as a closed opaque container into which 
stream the raw materials of photosynthesis, namely, 
carbon dioxide, light and water containing the neces- 
sary mineral elements. From this container are evolved 
the products of photosynthesis- oxygen gas and the 
reduced carbon compounds constituting the plant and 
its stored reserves. Heretofore, it has been possible to 
study in a quantitative way the nature of the process 
going on inside the opaque container only by varying 
external conditions and noting variations in the final 
products. Although there has been no serious doubt 
that the formation of sugar did not take place by the 
aggregation of six molecules of carbon dioxide, six 


ri5. XII. 1952] 

M. Calvin and P. >[assini: The I\itli u( Ciiboii in Pliolusyntheais. 


molecules of water and the requisite number of light 
quanta into a single unit followed by the rearrangement 
into hexose and molecular oxygen, no specific infor- 
mation was available as to the compounds which might 
act as intermediates. Assuming that such a chain of 
intermediates exists, it is quite clear that by setting up 
some photosynthetic organism, leaf or other suitable 
material, in a steady state of photosynthesis in which 
the various ingredients are being absorbed and pro- 
ducts formed in some uniform manner and injecting the 
labeled carbon dioxide into the entering carbon dioxide 
stream, we should find the label appearing successively 
in time in that chain of intermediates. This can be 
observed by stopping the entire process after a suitable 
lapse of time and examining the incorporated labeled 
carbon to determine the nature of the compounds into 
which it has been built. It is also clear that in addition to 
the identity and sequence of the compounds into which 
the carbon is incorporated, we may also determine the 
order in which the various carbon atoms within each 
compound acquire the label. With this type of infor- 
mation at hand it should be possible to reconstruct the 
sequence of events from the time of entry of the carbon 
atom into the plant as carbon dioxide until it appears 
in the various more or less finished products of the plant . 

Fig. 3.— .\lg.ic I'l.iiU. 

While photosynthetic experiments have been done 
with a vide variety of plant materials, the major ki- 
netic work has been carried out with suspensions of 
unicellular green algae. The reason for this lies in the 

I'ig. 4. — "Lollipop". 

fact that these algae may be obtained in a reproducible 
biological form relatively easily and in any amount. 
They are grown in the laboratory in a continuous cul- 
ture arrangement shown in Figure 3. The algae maybe 
harvested from these flasks daily or every other day, 
depending upon the type of material desired. Such 
cultures have been maintained in a continuous fashion 
over periods extending beyond several months. Most 
of our experiments have been performed with the uni- 
cellular green algae Chlorella or Scenedesmus. After 
harvesting the algae are washed with distilled water 
and resuspended in the medium in which the experi- 
ment is to be done. This suspension is placed in a flat 
vessel called a "lollipop", a photograph of which is 
shown in Figure 4. A stream of air containing carbon 
dioxide is passed through the algae while they are 
being illuminated so as to achieve a steady state of 

In order to begin the experiment the air stream is 
interrupted and the labeled bicarbonate is injected into 
the algal suspension. After the preselected period of 
time, the algae are killed by opening the large stopcock 
at the bottom of the flask, allowing the algal suspension 
to fall into alcohol in order to stop the reaction and 
extract the photosynthesized material. Although a 
variety of killing and extracting procedures have been 
tested, most of the experiments were performed by 
dropping the algae into alcohol so as to result in an 
80% alcohol solution. The total amount of carbon 
fixed is then determined by taking an aliquot of this 
entire suspension, evaporating it to dryness on an alu- 
minum disk and counting it on a Geiger counter'. The 
fraction soluble is determined by either filtering or 
centrifuging the suspension and then recounting the 
clear supernate or filtrate. 

The distribution of the fixed radiocarbon among the 
various compounds must now be determined. Since in 

' M. Calvin, C. HtlutLUERGER, J. C. Reid, Lt. M. Tolbert, and 
P. E. Yankwich, Iwtopic Carbon (John Wiley & Sons, Inc., New 
York, 1940). 




i . 



M. CvLViN and P. Massini: The Path of Carbon in Photosynthesis 


relatively short periods of time most of the fixed radio- 
activity is found in the soluble components, the prob- 
lem is one of analyzing for the distribution in the 
soluble fraction. This has been done by an application 
of the method of paper chromatography introduced 
and developed for amino acid analysis by Consden, 
Martin, and Synge'. It has since been applied to a 
wide variety of compounds and no detailed description 
of it will be given here. The unique extension to our 
work lies in the ability to locate particularly those 
compounds which contain the radioactive carbon atoms 
on the paper by means of a radioautograph of the 
resulting paper chromatogram obtained by allowing an 
X-ray film to remain in contact with the paper for a 
suitable period of time. Those areas of the paper which 
are occupied by radioactive compounds will, of course, 
expose the X-ray film. Such a map of the disposition 
of the radioactive compounds contained in an extract 
is shown in Figure 5. The chemical nature of the com- 
pounds defined by the exposed areas can be inferred 
from the position occupied by a compound with re- 
spects to the origin of the chromatogram. More precise 
determination of the chemical character is assisted by 
chemistry performed on the material eluted from the 
spot defined by the radiogram and rechromatography. 
Final identification, however, is usually dependent on 
the co-chromatography of the unknown, or questioned, 
radioactive material eluted from the paper with an 
authentic specimen of the suspected compound and 
the demonstration of the complete identity of the car- 
rier material as determined by some visible test on the 
paper with the pattern of radioactivity in the co-chro- 
matogram. The amount of radioactivity incorporated 
in these compounds can be determined quite accurately 
by using the X-ray film as a means of defining that 
area of the paper containing the compound, thus per- 
mitting the particular spot to be cut out from the larger 
and eluted from the paper and mounted on a plate to 
be counted. 

*f.^i=n8s»i:':r.--i' T.",i»W:' 



\ PHO*n(0«4.Y0tllATt 



mtUlOM mtOfHATE 

MRMTOM mOVMTl ft MMMOtC mOWlun' f 



•uftULOtc wmowtun • niiatB emioftnuTff 

Fig. 5, 

-Radiogram of a paper chromatogram from 10 s C**0, fixation 
ID the light by Scenedesmus. 

* R. Consden, A. H. Gordon, and A. J. P. Martin, Biochem. J. 
28, 224 (1944). - A. A. Benson, J. A. Bassham, M. Calvin, T. C. 
GooDALE, V. A. Haas, and W. Stepka, J. Am. Chem. Soc. 72, 1710 

















2° C 



























O onotphotc 











— • a — 



T8 U, 



Fig. 6, 

— Behavior of radioactivity in specific compounds in extracts 
Scenedestnus , exposed to radioactive carbon dioxide at 2*'C. 

A much simpler means would be to count the spot 
right on the paper with a Geiger counter. The fraction 
of the total amount of radioactivity in the spot which 
is thus registered by the Geiger counter is fairly con- 
stant for all compounds for any given chromato- 
graphic system. Thus, for most purposes it is sufficient 
simply to expose the paper to X-ray film in order to 
determine just where the radioactive spots are, and 
then having so defined them, to count them right on 
the paper for quantitative comparison, by the Geiger 
counter. It is clear from Figure 5 that the variety of 
products synthesized at room temperature by Scene- 
desmus (as well as by all other plants tried) is great, 
even in a very short time such as ten seconds. But even 
so, it is clear that the predominant compound as the 
time gets shorter is phosphoglyceric acid. 

This is even more strongly demonstrated when the 
experiment is carried out at reduced temperature, for 
instance 2°C, so as to slow down all of the reactions 
and enable us to see more clearly the earliest products. 
Figure 6 shows a plot of the concentration of radio- 
activity per unit of algae for three of the major early 
compounds, formed at 2°C. On such a plot as this, it 
is clear that those substances which are formed directly 
from carbon dioxide with no appreciable intermediates 


115. XII. 1952] 

M. Calvin and P. Massini: The Path of Carbon in Photosynthesis 


TabU I 
C^* Distribution in Photosynthetic Products of Barley and Scenedismus 


Glyceric Acid 

GlycoHc Acid 




— CHjOH 


— CHjOH 


C 2,5 

C 1,6 

Preillum : 
2 min dark . . . 

4 s PS 

15 s PS 

15 s PS 

30 s PS 


40 s PS 

60s PS 


5 s PS 

30 s PS 

30 s PS'MI' . . . 

60 s PS» 

60sPS«MI. . . . 

60 s PS» 

60s PS'MI. . . . 























50 ±5 




50 ±5 





' Experiments are steady-state photosynthesis, 10,000 footcandles unless otherwise stated. ' 1000 footcandles. ' Alanine obtained from this 
extract was 48% carboxyl-labeled. * Under the same conditions, Chlorella produced phosphoglycerate latelcd 93%, 3% and 2%, 
respectively. ' In this extract, malic acid was labeled 6*5% and aspartic acid 4% in the non-carboxyl carbons. ' 3000 footcandles. 

" Malonate inhibited. 

lying between them and carbon dioxide will be the only 
ones that will show a finite slope ; all others should start 
with a zero slope. A finite slope is certainly the case for 
phosphoglyceric acid and possibly for malic acid, in- 
dicating at least two independent catrbon dioxide fixing 
reactions, one leading to a three-carbon compound and 
the other producing a four-carbon compound'. 

Since the hexose phosphates appear extremely early 
in all of these photosynthesis experiments and because 
of the known close relationship between the hexose 
phosphates and phosphoglyceric acids in the glycolytic 
sequence, it seemed most reasonable to suppose that 
these hexose phosphates were formed from the phos- 
phoglyceric acid by a combination of the two three- 
carbon fragments derived from phosphoglyceric acid in 
an overall process very similar to, if not identical with, 
the reversal of glycolysis. 

One means of testing this suggestion would be a com- 
parison of the distribution of radioactivity in the three 
carbon atoms of glyceric acid with those in the hexose 
as shown in Table I. It thus appears that the hexose is 
indeed formed by the combination of two three-carbon 
molecules derived from the glyceric acid in such a 
manner that carbon atoms three and four of the hexose 
correspond to the carboxyl-carbon of the glyceric acid ; 
carbon atoms two and five with the alpha-carbon ; and 
carbon atoms one and six with the beta-carbon of the 

* E. J. Badin and M. Calvin, J. Am. Chem. Soc. 7^, 5266 (1950). 
- S. Kawaguchi, a. a. Benson, M.Calvin, and P. M. Hayes, J.Am. 
Chem. Soc. 7i. 4477 (1952). 

glyceric acid. This correspondence is maintained when 
the distribution in these two compounds (glyceric acid 
and hexose) is compared for a wide variety of different 

With this clear cut indication of the similarity be- 
tween the path of hexose synthesis and the known path 
of its breakdown, another means of testing how closely 
this parallelism might be followed suggests itself. The 
hexose derivative which is last in the sequence of 
changes prior to the breakdown of the carbon skeleton 
during glycolysis is the fructose-l,6-diphosphate. 
Correspondingly, then, it presumably would be the 
first hexose derivative to appear in the reverse direction. 
If this is the, case and, furthermore, if the hexose deriv- 
ative reservoirs involved in sucrose synthesis are more 
or less isolated from those involved in storage and gly- 
colysis, the radioactivity should appear in the fructose 
half of the sucrose molecule prior to its appearance in 
the glucose half. This is indeed the case'. However, 
sucrose does not seem to be formed by the simple re- 
versal of the sucrose phosphorylase system which was 
described for certain bacteria^, since for this to be the 
case, free fructose would have to be apparent in the 
photosynthesizing organism, whereats it is never so 
found, nor has the enzyme itself ever been isolated 
from any green plant. 

* S. Kawaguchi, A. A. Benson, N. Calvin, and P. M. Hayes, 
J. Am. Chem. Soc. 7i, 4477 (1952). 

' W. Z. Hassid, M. Doudoroff, and H. A. Barker, J. Am. Chem. 
Soc. se, 1416 (1944). - M. Doudoroff, H. A. Barker, and W. Z. 
Hassid, J. Biol. Chem. ISS, 725 (1947). 



M.C\uviN ami 1'. Massini: Tlie TjIIi ot Carbon in I'hotosyulhesis 


The recent identification' as uridine diphospho- 
glucose (U.D.P.G.) of the spot which had been previ- 
ously* called «the unknown glucose phosphate spot» 
has lead to another suggestion as to the mode of for- 
mation of sucrose. Glucose-labeled U.D.P.G. appears 
very early in the sequence of compounds formed. Fur- 
thermore, it has been possible to demonstrate the pres- 
ence in the hexose monophosphate area of a sucrose 
phosphate by using a carefully selected phosphatase, 
containing no invertase, in the treatment of this entire 
phosphate area'. We have suggested, therefore, that 
U.D.P.G. may be involved in sucrose synthesis in a 
manner similar to that of glucose-1-phosphate in the 
numerous phosphorylase reactions, with the difference, 
however, that the acceptor of the glucose moiety would 
be some phosphate of fructose, thus producing a sucrose 
phosphate. Recent work by Putnam and Hassid' gives 
further support to the idea that only phosphorylated 
derivatives of glucose and fructose are involved in 
sucrose synthesis in higher plants. They found that in 
sucrose synthesis, from labeled glucose in leaf punches, 
no free fructose was formed, although the sucrose be- 
comes equally labeled in both the glucose and fructose 
portions. Conversely, when labeled fructose is used, no 
free labeled glucose appears, while the sucrose is uni- 
formly labeled in both moieties. 

It is possible that compounds of the U.D.P.G. type 
could be concerned in the transformation of sugars and 
the subsequent incorporation into polysaccharides. 
Uridine diphosphate would thus serve as a carbon 
carrier in the same way that pyridine nucleotides and 
flavonucleotidcs are involved in hydrogen transfer ; the 
adenylic acid system in phosphate transfer ; and coen- 
zyme A in the transfer of acetyl groups. There is already 
some evidence for the existence of other members of 
the uridine diphosphate group from our own work, as 
well as that of others*. 

We may now turn our attention from the fate of the 
glyceric acid to the problem of its origin. An exami- 
nation of Table I indicates quite clearly that the first 
position in the glyceric acid to become labeled is the 
carboxyl group. As time proceeds, the other two carbon 
atoms in the glyceric acid acquire radioactivity and it 

' J. G. Ulchana.n €t at., in press. - J. G. Blchanan, J. A. Uass- 
HAM, A. A. UtNSON, D. F. Bradley, M. Calvin, L. L. Dals, M. 
Goodman, P. M. Hayls, V. H. Lynch, L. T. Norris, and A. T. 
Wilson, Phosphorus Metabolism, \'ol. II (Johns Hopkins Press, 
Baltimore, Maryland. 1952), in press. 

' S. Kawaclchi, a. a. Benson, N. Calvin, and P. M. Hayls, 
J. Am. Chem. Sor. 71, AV7 (1052). 

' J. G. Buchanan, J. A. Bassham, A. A. HtNSON, V. V. Uradllv, 
.M. Calvin, L. L. Dals, M. Goodman, P. M. Hayes, V. H. Lynch, 
L. T. Norris, and A. T. Wilson, Phospliorus Metabotism, \'ol. II 
(Johns Hopkins Press, Baltimore, Maryland, 1902), in press. - J. G. 
Buchanan, in press. 

• E. W. Putnam, Thesis (University of California, Berkeley. l'J5'.>). 

* R. Caputto, L. F. Leloir, C. E. Cardini, and A. C. Paladim. 
J. Biol. Chem. IS4, 333 (1950). - A. C. Paladini and L. F. Leloir, 
Biochem. J. 51, 126 (1951). - J. T. Park, J. Biol. Chem. 1S4, 885 

appears that they acquire it at equal rates, at least 
within the present accuracy of the experiments. 

It thus appears that the most rapid reaction which 
carbon dioxide can undergo at least at high light in- 
tensities, is a condensation with a Cj fragment leading 
directly to phosphoglyceric acid. An examination of the 
chromatograms of a very short photosynthetic period 
shows glycine and glycolic acid as the only two-carbon 
compounds present. The distribution of radioactivity 
among the carbon atoms of these two compounds is 
always equal and the same and corresponds very well 
with that in the alpha- and beta-carbon atoms of the 
glyceric acid, as may be seen from Table I. This sug- 
gests that glycolic acid either is in the direct line for 
the formation of the Q carbon dioxide acceptor, or is 
very closely related thereto. 

The question now arises as to the source of this Cj 
carbon dioxide acceptor. There are, of course, only two 
possibilities for its origin. Either it results from a one- 
plus-one combination or it must result from the split- 
ting of a four-carbon compound or a larger one. In order 
for it to result from the combination of two one-carbon 
fragments there must exist as an intermediate some 
one-carbon compound more reduced than carbon diox- 
ide which, in turn, may combine either with itself or 
with carbon dioxide. Furthermore, the reservoir of 
this one-carbon intermediate would have to be vanish- 
ingly small since all attempts to find labeled, reduced, 
one-carbon compounds, such as formic acid or formal- 
dehyde, in the early stages of photosynthesis have 
failed and, in addition, the resulting two-carbon frag- 
ment is very nearly equally labeled in both carbon atoms. 
One would also expect that these one-carbon com- 
pounds would tend to disappear under conditions of 
low carbon dioxide concentrations leading to the disap- 
pearance of the two-carbon condensation product re- 
sulting from them. This leads us to the supposition that 
the formation of glycolic acid would be expected to 
drop off under conditions of low carbon dioxide con- 
centration which is the reverse of what is observed. 

We are thus left with the following possibility for the 
C'a compound -the cleavage of some C4 or larger struc- 
ture. The fact of the early appearance of label in malic 
acid, taken together with the lack of any appreciable 
amounts of label in the compounds of the tricarboxylic 
acid cycle', led us to the supposition that malic acid 
was either a precursor to, or very closely related to, a 
four-carbon compound which could be split to produce 
the required two-carbon fragment. 

In the course of the search for the two-carbon ac- 
ceptor, and its immediate precursors, two new com- 
pounds were identified as early products of carbon 
dioxide incorporation which seem to have little to do 
with the direct synthesis of hexoses and, therefore, had 
a very likely function in the regeneration of the two- 

1 A. A. Benson and M. Calvin, J. Exptl. Botany ;, 63 (1950). 


[15. XII. 1952) 

M. Calvin and P. Massini: The Path of Carbon in Photosynthesis 


carbon acceptor. These were the phosphates of the 
seven carbon sugar sedoheptulose and of the five car- 
bon sugars ribulose, ribose and arabinose'. 

The question immediately presents itself as to the 
relation between these two compounds along the path 
of carbon assimilation, not only with each other but 
with the precursors which are already known and the 
possible products that might be formed from them. 
The attempt to answer this question focuses our at- 
tention once again upon some of the shortcomings and 
limitations of the method of observation that we are 
using and the nature of the exjjeriment which we are 
performing. Our initial hope of determining the se- 
quence of intermediates by a simple observation of a 
sequence of compounds into which radioactivity has 
been incorporated in steady state experiments is now 
complicated by the uncertainty as to the amount of 
compound present during the steady state. It is easy 
to visualize a situation in which the actual amount of 
intermediate present during the steady state is so small 
as to escape observation by our methods, or perhaps 
even to be so unstable as to be lost by our methods of 
observation. This complete failure of a compound to 
appear on a chromatogram, although it might con- 
ceivably be an intermediate, is, of course, an extreme 
case. The more usual situation is one in which most of 
the intermediates are present but in varying concen- 
trations in the steady state. Under such conditions a 
single or even several observations of the relative 
amount of radioactivity incorporated into a variety of 
compounds would not necessarily be any real criterion 
of the relative order of these compounds in the se- 
quence of events. 

In order to achieve the full value of the method of 
observation then, it becomes necessary to perform 
rather extended kinetic experiments in which the ap- 
pearance of radioactivity in all compounds is plotted 
as a function of time at sufficiently short intervals to 
enable a rather accurate and detailed curve to be 
obtained. Furthermore, the distribution of radioacti- 
vity among the atoms within each compound should 
also be determined as a function of time. The validity 
of any proposed sequence of events could then be de- 
termined by a comparison of the calculated appearance 
and distribution curves with those actually observed. 
In order to calculate such appearance curves, as well 
as the distribution curves amongst the atoms in each 
compound, one can set up a system of linear differen- 
tial equations based upon the following model ; 


-+ B 


where COj represents the entering carbon dioxide; 
.4, B, etc. represent intermediates involved in carbon 

' A. A. Bt.MSO.s, J. A. BASSIIA.M, .M. Calvin, A. G. Hall, H. E. 
HiRSCii, S. Kawaguchi, V. H. Lynch, and N. E. Tolbert, J. Biol. 
Chcm. 196, 703 (1952). 

dioxide assimilation; S represents more or less final 
storage product; /? is a measure of the total rate of 
carbon dioxide assimilation in the steady state ex- 
pressed in moles of carbon per minute. 

The rate of change of the specific activity of a single 
carbon atom in A , given by X^ , is then expressed by 
Equation (2). (The specific activity of the entering 
carbon dioxide is here taken as unity. [A], the concen- 
tration of the compound A, is independent of time.) 

^ = w (1-^^)- 


The specific activity of the corresponding atom in 
compound B is given by an exactly similar Equation 

dXo R 





Equations of identical form may be written for every 
atom of every compound that might be considered an 
intermediate. These equations may be solved expli- 
citly by means of a differential analyzer provided two 
parameters are known. These are the total rate of entry 
of carbon into the system during the steady state, R, 
and the steady state concentration of each atom which 
might be considered as lying along the path of carbon 
assimilation [,4], [B], etc. 

It is clear that if such compounds (whose prime func- 
tion it is to serve as carbon carriers between the en- 
tering carbon dioxide and the final storage products in 
the plant) do indeed exist in biological systems they 
would very soon become saturated with radioactivity. 
By this is meant that the amount of radioactivity ob- 
served in that particular compound would very soon 
reach a maximum value and remain that way. The 
reason for this is that by definition the amount of these 
intermediate compounds is not changing, and also is 
small compared to the total amount of carbon the plant 
assimilates during the experiment. Since all of the 
carbon, or at least most of it, must pass through these 
reservoirs of intermediates they will very soon acquire 
the same specific activity as the entering carbon di- 
oxide. In contrast to this, those materials which are 
not functioning as simple intermediates but rather are 
functioning as storage reservoirs, or are very distant 
from the immediate photosynthetic intermediates, will 
not acquire radioactivity as rapidly, or if they do they 
will not become saturated as rapidly as those which are 
directly mvolved in the path of carbon assimilation. 
The amount of radioactivity found in those compounds 
which saturate in a relatively short time now provides 
a relatively easy method of determining the size of the 
functioning reservoirs of these compounds which are 
directly engaged in the path of carbon assimilation. A 
simple measurement of this amount compared to the 
specific activity of the entering carbon dioxide will 
provide a measure, in moles per unit volume of the 



M. Calvin and P. Massini: The Path of Carbon in Photosynthesis 


biological material, of the compound in question. 
Furthermore, having once achieved a relatively uni- 
form label in these photosynthetic intermediates, it 
becomes possible to follow the behavior of the reservoir 
size as a function of change in external variables, for 
example, light intensity. We have chosen to include in 
this review a more or less detailed description of just 
this determination of the effect of light intensity upon 
reservoir sizes as a means of describing the general ex- 
perimental technique which is involved. 

Steady state and reservoir sizes— Methods and results 

The apparatus used for these experiments was con- 
structed to permit the algal suspension to be left under 
controlled external conditions (illumination intensity, 
temperature, carbon dioxide and oxygen concentration) 
while photosynthesizing for at least one hour. Further- 
more, it was required that the change, natural to radio- 
active carbon dioxide, which was to be circulated in a 
closed system, and the withdrawal of several samples 
at given time intervals be accomplished with a mini- 
mum of change in these conditions. 

The apparatus consisted of : 

(a) A square illumination vessel A (Fig. 7) made out 
of Lucite (polyacrylic plastic), 49 cm high, 11 cm wide 
and 0-7 cm thick (inside dimensions). The bottom was 
provided with a gas inlet tube with five small holes to 
allow good contact between gas and liquid and a drain 
tube closed with a screw clamp. The top of the vessel 
was provided with a gas outlet tube. A water-alcohol 
mixture from a constant temperature bath was allowed 
to flow over the outer surfaces of the vessel in order to 
control the temperature of the suspension. 

Fig. 7.— Diagram of the assembly for steady state photosynthesis. 
(For explanation of the letters, see text.) 

[b) Two illumination banks (represented by B), each 
with four fluorescent tubes (General Electric, quality 
white, 20 W each), providing an almost uniform illumi- 
nation over the whole surface of the vessel, of 7 x 10* 
ergs. /cm ^ (roughly 700 footcandles). 

Kig. S.— .\sscnibly for steady state photosynthesis. (For explanation 
of the letters, sec text.) 

(c) An ionization chamber C, connected to a record- 
ing vibrating reed electrometer, to record the activity 
of the gas leaving the vessel continually during the run. 

{d) Three gas traps D, to permit the addition of a 
known amount of radioactive carbon dioxide to the 
system, and trap the remaining radioactivity after 
the run. 

(e) A flask E, of 5 1 volume, containing a mixture of 
1% radioactive carbon dioxide in air. The reservoir 
contained so much carbon dioxide that the algae assi- 
milated no more than 20% of it during a run. 

(/) A gas circulating pump F of the rubber tubing 
type, and a flow meter G. 

(g) A system of four-way stopcocks H, which per- 
mitted the vessel to be flushed with a mixture of 1 % 
ordinary carbon dioxide in air, from the cylinder I. The 
assembly is shown in Figure 8. 

In a typical experiment, 2 cm' (wet packed) of one- 
day old Scenedesmus, washed and resuspended in 200 
cm' of deionized water, were placed in the vessel and 
aerated with the ordinary gas mixture for at least one- 
half hour, while the mixture of radioactive carbon 
dioxide circulated in the gas system for thorough 
mixing, without passing through the vessel. The sus- 
pension was kept at 24 °C. After this time, during which 
a steady state of photosynthesis had been reached, the 
radioactive mixture was passed through the vessel in 
place of the ordinary gas mixture, by a manipulation 


[13, XII. 1952J 

M. Calvin and P. Massini: The Path of Carbon in Photosynthesis 



Fig. 9. — Radiogram of a paper chromatograra from lu min. 0^*0^ 
fixation in light by ScCTi^iitfSmui. I % suspension, 1% Ct)j in air; light 
intensity 7 x 10* ergs./cm^-s. D ll.A.P. ;dihydroxyacetone phosphate; 
P.E.B. :phosphoeno]pyruvir acid; P.M.P.:pentose monophosphates; 
P.Go.A.:phosphoglycolicacid: P.G.A.:phosphoglycericacid; H.M.P.: 
hexose monophosphates; U.P.rpentose and hexose diphosphates. 

of the pair of stopcocks at H, and samples of 20 cm' of 
the suspension withdrawn at intervals of five or ten 
minutes. These samples were dropped into 80 cm' of 
alcohol of room temperature, to make an e.xtraction 
in 80% alcohol. After 30 min of photosynthesis, the 

lights were turned off and the suspension allowed to 
remain in the dark for a period of 5 min, during which 
time again several samples were withdrawn, and treated 
in the same manner. In one experiment another light 
jseriod followed the dark period. 

The samples were shaken for 1 h and centrifuged. 
The residue was re-extracted in 50 cm' of 20% alcohol 
at room temperature, centrifuged, and re-extracted 
again with 20 cm' of water. The extracts were concen- 
trated together to 0-5 cm'. 

An aliquot of the concentrate equivalent to 30 /il of 
packed cells was evaporated on a corner of a filter paper 
(Whatman #1), and the chromatogram run with water- 
saturated phenol in one direction and n-butanol- 
propionic acid-water in the other. The chromatograms 
were exposed to X-ray film for about two weeks'. The 
labeled compounds appeared on it as black spots. 
Figure 9 shows the radiogram for ten minute photo- 
synthesis of Scenedesmus. The amount of radioactivity 
contained in the different compounds was determined 
by counting the corresponding spots on the paper di- 
rectly with a large-area Geiger-MCller tube with 
thin mica window. The compounds were identified by 
a combination of the following criteria: (a) Their posi- 
tion on the paper; (b) the spot was cut out, eluted from 
the paper with water and run again in suitable solvents, 
together with such an amount of the suspected com- 



Fig. 10. — C'*0, fixation by Scenedesmus. 1 % suspension, 1 % COg in 
air, light intensity 7 x 10* ergs./cm*-s. 

TIME (Kiln.) OF EXPOSURE TO (?*0j 

Fig. 11.— Behavior of radioactivity in specific compounds in the ex- 
tract from the experiment of Figure 10. 

pound that it could be detected by a specific spraying 
reagent. The black spot on the film had to coincide 
accurately with the color reaction ; (c) the eluted spot 
was chemically transformed {e.g. treating the sugar 
phosphates with phosphatase) and the resulting com- 
pound cochromatographed with carrier detectable by 

Figure 10 shows the total and the extracted amounts 
of radiiKarbon fixed by 1 cm' cells during 30 min of 

' M. Calvin (. Chem. Education ?6, 639 (1949). 



M. Calvin iiud P. M.\ssl^•l: Tlic Path of Carbon iu Photosynthesis 





labU 11 

Steady State Concentrations of Some Compounds Involved iu the 

Photosynthesis Cycle. Scenedesmus, experimental conditions as in 

Figure Iu 


Fig. 12. — Behavior of radioactivity in specific compounds in the ex- 
tract from an experiment done under conditions corresponding to 
those of Figure 10. 

photosynthesis followed by 5 min of darkness. The 
slope in the total fixation curve in the light corres- 
ponds to a 13 cm^ COj assimilation (N.T.P.) per hour. 

Figure 11 shows the amount of radioactivity incor- 
porated into sucrose and three phosphorus compounds 
for the experiment of Figure 10. 

Figure 12 gives the number of counts in sucrose, 
glutamic, malic and citric acid, for a different experi- 
ment of 15 min photosynthesis, followed by 10 min 
dark, and again 5 min of photosynthesis. 

Although the variation between experiments is quite 
high, there are some striking features which are com- 
mon to all : 

(1) The curves of some of the compounds show a 
marked decrease in slope after 5 min of photo- 
synthesis. This quite clearly indicates the presence of 
rapidly turning-over reservoirs in the photosynthesis 
cycle which are then thoroughly labeled and reach the 
specific activity of the fed carbon dioxide : Diphosphate 
area (mainly ribulose diphosphate) ; hexose-monophos- 
phate area (50% glucose-, 26% sedoheptulose-, some 
fructose- and mannose-monophosphate) ; phospho- 
glyceric acid. The leveling off of these curves permits 
the calculation of the concentration of the reservoirs of 
those compounds in the photosynthesis cycle, by divid- 
ing the measured amount of radioactivity per carbon 
atom by the specific activity of the fed carbon dioxide'. 

Table II gives the steady state concentrations during 
photosynthesis for some compounds determined by 
this method. 

(2) The fact that the activity vs. time curves show 
a definite yet low slope for as long as 30 min can be 
taken to indicate that the breakdown of carbohydrates 


jumoles/cni' cells ^ 

Phosphoglyceric acid 

Dihydroxyacetone phosphate . . 

Fructose phosphate 

Glucose phosphate 

Mannose phosphate 

Sedoheptulose phosphate .... 
Ribulose diphosphate 









continues throughout the illumination, i.e. their for- 
mation from photosynthetic intermediates is revers- 
ible. Thus, there are two sources of the intermediates: 
{a) the carbon dioxide fed; the amount of compound 
formed from this source reaches the maximum specific 
activity in 5 to 10 min ; (6) the carbohydrate pool of the 
ceUs; the amount formed from this source is labeled 
only slowly since the specific activity of the carbo- 
hydrate pool rises slowly due to the large size of the pool . 
(3) Other compounds show almost constant rate of 
labeling during the whole period of photosynthesis; 
sucrose, malic and glutamic acid. For this and other 
reasons it is clear that these compounds are not in the 
photosynthesis cycle, but are formed during the photo- 
sjmthesis at a constant rate. Their large reservoirs in 
the cells are labeled only slowly. 

Table 111 

Phosphatase Treatment of H.M.P. Area after au min Photosynthesis 
and 30 min Photosynthesis Followed by 5 min Dark 


Number of counts/min 
on paper 

30 min P.S. 

30 min P.S. 
5 min D. 









(4) When illumination is interrupted there appears 
a sudden great increase in the concentration of phos- 
phoglyceric acid (followed by a slow decrease after 2 
min), and an almost complete depletion of the diphos- 
phate area. Analysis of the monophosphate area 
showed that the amount of sedoheptulose phosphate 
decreased also (cf. Table III). The concentration of 

' The efficiency factor of the counting of spots on papers has been 
determined by converting three cut out spots to barium carbonate 
and measuring their activity in an ionization chamber. It is 19 
disintegrations per count. 

^ Volume measured as wet packed cells 

- An appreciable fraction of this count is certainly hexosc so that 
one may estimate a maximum value ot the heptose at around 800 


115. XII. 19521 

M. Calvin and P. Massini: The Path of Carbon in Photosynthesis 


malic acid decreases as well. The rate of labeling of 
glutarriic acid is increased greatly after a short induc- 
tion period; citric acid, which contains little activity 
during the whole light period, shows a sudden increase 
in the dark, followed by a slow decrease. The labeling 
of sucrose continues at the same rate as in light for 
about 1 min, after which it is stopped almost com- 

Both experiments gave the same picture for most of 
the compounds, with the two exceptions: In the second 
experiment the diphosphate area, which in the first 
contained almost the same number of counts as phos- 
phoglyceric acid during the light, had only about 15% 
of it in this second run. This value dropped to 5% in 
the dark. The phosphoglyceric acid showed a hardly 
significant rise in the dark during the first 2 min, but 
again a slow decrease after 5 min. Although we do not 
know why in this experiment the concentration of 
ribulose diphosphate was so low in the light, the co- 
incidence with the lack of increase of phosphoglyceric 
acid points to a connection between both effects. 

(5) In the light following the dark, the diphosphates, 
phosphoglyceric and malic acid increase again. 

^'^ "J^f GLUTAMIC 

60sL-60sO^^HHH^M^^^^H 37 


l20sL ^^Hi 10 




60tL-60sO ^^^BH 
I20SL ^^8 2 


■^^■1 '*3 


eosi -eosD ^^^^m 

^1 ^?7 

I90O ^^^^m 

^^H 37 3 

60sL-60sO I 

I 91 



After one-half hour, the aeration bubbler was taken 
out and a suitable amount of radioactive bicarbonate 
(sodium) solution added. (The algae, which were grown 
in slightly acid medium, had enough buffering capacit\- 
to convert the bicarbonate to carbon dioxide). The 
vessel was immediately stoppered and shaken in the 
light. .After 1 min. the suspension was drained into 


I-'ig. 13.— Effect of light and dark on the labelingof glutamic and citric 

acid. 0-1% suspension, light intensity 1-6 < 10' ergs/cm'-s (Numbers: 

counts/min x 10~^ on paper per cui' cells). 

The effect of dark on the labeling of glutamic and 
citric acid was already reported in an earlier paper^ 
and studied more closely in the following experiments: 
0-2 cm' wet packed algae (ChloreUa pyrenoidosa) were 
suspended in 200 cm' distilled water, illuminated in a 
flat circular vessel of 1 cm thickness by incandescent 
lights through an infrared filter (intensity 1 -6 x lO^ergs/ 
cm^-s) and aerated with 008% carbon dioxide in air. 
The low concentration of cells was chosen to avoid shad- 
ing of cells in the suspension, so that during the light 
period all the cells were illuminated continually. 

• .\I. Calvin, J. Chem. Education SC, 639 (1949). 

a darkened flask, and after another minute poured into 
four times its volume of boiling alcohol. Control samples 
were treated in the same way, but kept in the light, in 
contact with radioactive carbon dioxide for 1 and 2 
min, respectively. The analysis of the fixed radioacti- 
vity was performed by paper chromatography and 
radioautography with the technique already described. 
The results are shown in Figure 13. 


It has already been pointed out that photosynthesis 
is not a mere reversal of respiration ; this was supported 
by the observation that the carbon of newly formed 
photosynthetic intermediates is not available for res- 
piration while the light is on'. We may thus represent 
the relationship between photosynthesis and respira- 
tion by the following scheme (See Figure 14). The 
labeling of the Krebs cycle intermediates through the 
storage products (carbohydrates, fats, proteins) of the 
cells is a slow process, due to the relatively large size 
of the storage pools. The fact that the photosynthesis 
intermediates find their way into the tricarboxylic acid 
cycle very rapidly after the light is switched off means 
that there is another connection between the two 
cycles which is blocked as long as the light is on but 
becomes accessible in the dark. This was interpreted in 
earlier work^ in terms of the action of the light in 
maintaining at low concentration the intermediate re- 
quired for entry into the tricarboxylic acid cycle. A 
closer specification of how this is accomplished is now 
possible since the discovery that alpha-lipoic acid is a 

' M. Calvin, J. Chem. Education J«, 639 (1949). - J. W. Weicl, 
P. iM. Warrington, and M. Calvin, J. Am. Chcm. See. 73, .lO.^ 

2 M. Calvin, J. Chom. Education J6, 630 (1949). 



M. Calvin and P. Massini: The Path of Carbon in Photosynthesis 

[ExpebientiaVol.VI 11/12) 

cofactor for the oxidative decarboxylation of pyruvic 
acid to an active acetyl group' which is the one reaction 
known to feed the Krebs cycle'. The mechanism of the 
reaction may be written this way: 


CH, HC - CH,-CH,-CH,-CO-Thiamin+ CH,-CO-COOH 

I I (Co-pyruvate oxidase) (Pyruvic acid) 


/ \ 
CH, CH- 

Cocnzymc A 





CH, CH-ff+ Acetyl CoA+ CO, 

Tlie reduced lipoic acid complex would then be reoxi- 
dized to the disulfide form by a suitable oxidant (e.g. 
pyridine or flavin nucleotides). In order that the oxi- 
dation of pyruvic acid can proceed, the enzyme has to 
be present in its oxidized form. If it is kept in its reduced 
form under the influence of the light-produced reducing 
f)ower, the reaction cannot proceed and the pyruvic 
acid formed during photosynthesis will not find its way 
into the respiratory cycle. The reaction is inhibited 
because only a small amount of the enzyme catalyzing 
it exists in the required form, most of it being kept in 
the other form under the "pressure" of the reducing 
power generated by the light energy. This recalls a 
similar phenomenon which has been known for a long 
time, i.e. the suppression of the fermentation of carbo- 
hydrates in favor of their oxidation under aerobic con- 
ditions (Pastruk effect). This effect has been explained 
in a manner similar to the one used here to account for 
the inhibition of the respiration of photosynthetic in- 
termediates'. The reduction of acetaldehyde to alcohol 
requires a dehydrogenase in its reduced form; under 
aerobic conditions the dehj'drogenase exists primarily 
in its oxidized form, and the acetaldehyde instead of 
beijig reduced is oxidized to acetic acid. 

The sudden rise in phosphoglyceric acid and the 
decrease in ribulose diphosjihate and sedoheptulose 

' L. J. Rfed, I. C. Cunsalus, et at.. J. Am. Chem. Soc. ?J, 5920 
(1951). -E. L. Patterson, rf a/., J. Am. Chem. Soc. 7J, 5919 (1951|. 
- I. C. GuNSALUS, I.. Struclia, and U. I. O Kane, ,1. Biol. Cheni. 
J9<, 859 (1952).- L. J. Reed and B. G. DeBusk, J. Am. Chem. Soc. 
r<, 3457 (1952). -M. W. Bullock, <( a/., J. Am. Chem. Soc. 7<, 3455 

* S. OcpcoA, J. R. Stern, and M. C. Schm idfr. J. Biol. Chem. 
/9J. 691 (1951). - S. KoRKEs, A.DelCamillo, I.C.Gvnsalus, and 
S. OCHOA, J. Biol. Chem. /93, 721 (1951). 

• O. Meverhof, Amer. Scientist iO, 483 (1952). 

phosphate in the dark period, together with the obser- 
vation that the dark rise in phosphoglyceric acid is 
absent when the ribulose diphosphate concentration 
was low during the light, confirms the earlier suggestion 
that the phosphates of the C, and Cj sugars are pre- 
cursors of the C, carbon dioxide acceptor*. ThLs, togeth- 
er with evidence gathered in previous work* leads to 
the following scheme for the photosynthetic cycle' 
(Fig. 15). 

Upon this basis an attempt might be made to relate 
the two effects as follows ; when the light is turned off, 
the reduction reactions requiring light are stopped, 
whereas cleavage and carboxylation reactions continue 
until their substrates are exhausted. Presumably, this 
would lead to a depletion of the Cj and C, sugars, the 
synthesis of which requires reduction steps (particu- 
larly the six-equivalents leading to the tetrose which 
itself is a very small reservoir), and a rise of phospho- 
glyceric acid, the further fate of which is also dej)endent 
upon reduction. However, a number of arguments seem 
to contradict this view : (I) The observation that plants 
fix radiocarbon in the dark immediately following a light 
period at low carbon dioxide concentration, to form 
a similar pattern of compounds as the one found in 
photosynthesis shows that the sequence following phos- 
phoglyceric acid is not blocked at once upon cessation 
of illumination, but that the cells contain sufficient 
reducing power to transform some phosphoglyceric 
acid intocarbohydrates , (2) the cleavageof the pentoses 
and heptoses into the Cj carbon dioxide acceptor and 
a triose and pentose respectively is dependent on a 
reduction step as well. 

Fig. 15. 

We are thus led to the suggestion that the rise in 
phosphoglyceric acid is not be explained by a mere 
interruption of the sequence, but that the rate of pro- 
duction of phosphoglyceric acid at some time in the 

' A. A. Benson, J. A. Bassham, M. Calvin, A. G. Hall, H. E. 
HiRscH, S. Kawaguchi, V. H. Lvnch, and N. E. Tolbert, J. Biol. 
Chem. ;9(i, 703 (1932). 

' S. Kawaguchi, A. A. Benson, M. Calvin, and P. M. Mayes, 
J. Am. Chem. Soc. 7/, 4477 (1952). - M.Calvin, The Harvey 
Lectures 46, 213-251, 1951, in press. 

' This scheme is intended to represent only changes in the carbon 
skeletons. The reducing equivalents are indicated only to show redox 
relationships between the known compounds. A number of the 
isolated compounds are isoxiraers and have not been included. 


[15. XII. 1952] 

M. Calvin and P. Massini: The Path of Carbon in Photosynthesis 



CO,. 2[H] 




phosphoglyceric acid 

^^^^ phosphoglyceric acid 

first minute of darkness is actually higher than it is in 
the steady state photosynthesis. This would be the case 
if the C3-C2 cleavage of ribulose diphosphate, which in 
photosynthesis presumably yields a triose phosphate 
molecule beside the Q carbon dioxide acceptor, in the 
dark yelds a molecule of phosphoglyceric acid instead 
of the triose molecule. The overall reactions may be 
represented above (not a mechanism). 

This hypothesis is supported by the fact that the 
triose phosphate also decreases in the dark. 

The fact that the net result of the reaction sequence 
in the light from ribulose diphosphate to phospho- 
glyceric acid and triose phosphate is a reductive car- 
boxylation and thus the reversal of the oxidative 
decarboxylation which, in the case of pyruvic acid, 
requires the presence of a cyclic disulfide compound 
leads to the idea that the former sequence might be 
catalyzed by a similar enzyme. This idea seems to be 
supported by the results of an experiment performed 
in this laboratory some time ago, which were difficult 
to explain'. 

In order to examine the relation between photosyn- 
thesis and the glycolytic cycle, a series of experiments 
similar to those described previously were performed 
with added iodoacetamide which is known to inhibit 
the action of triose phosphate dehydrogenase', pre- 
sumably through a reaction with its sulfhydryl group'. 
A 1% suspension of Chlorella in phosphate buffer was 
allowed to photosynthesize in light of 2500 footcandles 
and an atmosphere of 1% carbon dioxide, 5% oxygen 
and 94% nitrogen. At various times before adding the 
radioactive bicarbonate solution, iodoacetamide was 
added to give a 1-5 x 10"* M solution. 1 min after 
adding the radiocarbon, the cells were killed and 

After 8 min contact with iodoacetamide, the cells 
were still able to fix 75% as much carbon dioxide as 
non-poisoned cells otherwise treated the same way 
(control). The amount of radioactivity in phospho- 
glyceric acid was 50% of the control, and the amount 
in sucrose had reached a sharp maximum of 3-5 times 

' W. Stepka, Thesis University of California (June 1951). 

» O. Meyerhof and W. Kiessling, Biochem. Z. 28/, 249 (1053). 

' I.. Rapkins. C. r. See. Biol. (Paris) HJ, 1294 (1933), 

that in the control. There was practically no radio- 
activity in the ribulose diphosphate. After 90 min of 
exposure to the poison the cells had practically lost 
their ability of photosynthesis. 

If, in the proposed photosynthetic cycle, the cleav- 
age of the heptose and pentose phosphates is depen- 
dent on an enzyme containing sulfhydryl groups, which 
were more sensitive to iodoacetamide than the triose 
phosphate dehydrogenase, a picture similar to the one 
described would be expected : After short exposure to 
the poison, in relatively low concentration, the lack of 
Cj carbon dioxide acceptor would slow down the photo- 
synthetic cycle. The synthesis of carbohydrates, how- 
ever, would proceed almost without inhibition, thus 
decreasing the concentrations of the intermediates in 
the cycle. This would allow the compounds to reach a 
higher specific activity during the period of exposure 
to radiocarbon (cf. equation (2), change of specific 
activity inversly proportional to concentration]. At 
some time after administration of the poison, the su- 
crose would be labeled faster than in the control due to 
the higher specific activity of its precursors. After a 
longer period, however, the rate of synthesis of sucrose 
would decrease because the pool of its precursors would 
be exhausted. 


Die Trennung des Phanomens der Photosynthese 
griiner Pflanzen in eine Lichtreaktion und die vom Licht 
unabhangige Reduktion der Kohlensaure warden di.s- 

Die Reduktion der Kohlensaure und das Schicksal des 
assimilierten Kohlenstoffs wurden untersucht mit Hilfe 
der Spurenmethode (Markierung der assimilierten Koh- 
lensaure mit C") und der Papierchromatographie. Ein 
Reaktionszyklus wird vorgeschlagen, in dem Phosphogly- 
zerinsaure das erste isolierbare Assimilationsprodukt ist. 

Analysierung des Extraktes von Algen, die in einem 
stationaren Zustand fiir langere Zeit radioaktive Kohlen- 
saure assimilierten. lieferte weitere Auskunft iiber den 
vorgeschlagenen Zyklus und gestattete, die am Zyklus 
beteiligten Mengen einiger Substanzen ungefahr zu be- 
stimmen. Die friihere Vermutung. dass Licht den Res- 
pirationszyklus beeinflusst, wird bestatigt. Die Moglich- 
keit der Mitwirkung von a-Liponsaure (a-lipoic acid) oder 
einer verwandten Substanz, bei diesem Effekt und im 
Photosynthesezyklus, wird erortert. 


[Reprinted from the Journal of the American Chemical Society. 76, 1760 {1954).) 
CopyriKht lfl54 by the American Chemical Society and reprinted by permission of the copyright owner. 

[Contribution from Radiation Laboratory and Department of Chemistry, University of 

California, Berkeley) 

The Path of Carbon in Photosynthesis. XXI. The Cyclic Regeneration of Carbon 

Dioxide Acceptor^ 

By J. A. Bassham, A. A. Benson, Lorel D. Kay, Anne Z. Harris, A. T. Wilson and M. Calvin 

Received October 16, 1953 

Photosynthesizing plants have been exposed to C'Oj for short periods of time (0.4 to 15 sec.) and the products of carbon 
dioxide reduction analyzed by paper chromatography and radioautography. Methods have been developed for the degra- 
dation of ribulose and sedoheptulose. These sugars, obtained as their phosphate esters from the above C'»Oj exposures and 
from other experiments, have been degraded and their distribution of radiocarbon determined. The distribution of radiocar- 
bon in these sugars, and other data, indicate that sedoheptulose phosphate and ribulose diphosphates are formed during 
photosynthesis from triose and hexose phosphates, the latter being synthesized, in turn, by the reduction of 3 phosphoglyceric 
acid. Further evidence has been found for the previously proposed carboxylation of ribulose diphosphate to phosphoglyceric 
acid. Free energy calculations indicate this step would proceed spontaneously if enzymatically catalyzed. The efficiency 
of this cycle for reduction of CO2 to hexose would be 0.9 if the reduction of each molecule of PGA requires the concurrent 
conversion of one molecule of ATP and one of DPN (red) to ADP, inorganic phosphate and DPN (ox.). 

Previously reported tracer studies of the path of action leading to phosphoglyceric acid (PGA)' 
carbon in photosynthesis' led to the conclusion which is then reduced and condensed to fructose 
that carbon is incorporated by a carboxylation re- (3, ^-he following abbreviations win be used throughout this paper: 

(1) The work described in this paper was sponsored by the U. S. PGA, phosphoglyceric acid; DHAP, dihydroxyacetone phosphate; 
Atomic Energy Commission. This paper was presented before the FMP, fructose monophosphate; GMP, glucose monophosphate; 
Division of Biological Chemistry. American Chemical Society, at the SMP, sedoheptulose monophosphate; RDP ribulose diphosphate; 
124th National Meeting. Chicago. Illinois. September. 19.53. ADP, adenosine diphosphate. ATP adenosine triphosphate; DPN, 

(2) M Calvin. "The Harvey Lectures," Charles C Thomas Pub- diphosphopyridine nucleotide (Coenzyme I), oxidised form; DPNlHi], 
lisbing Company. Spring&eld, 111., 1050-61, p. 218. diphosphopyridine nucleotide, reduced form. 


April 5, 1954 

Cyclic Regeneration of Carbon Dioxide Acceptor 


and glucose phosphates by a series of reactions simi- 
lar to a reversal of glycolysis. These conclusions 
were supported by the observations that when car- 
bon- 14 is administered to the photosynthesizing 
plant as C'Oj, the first radioactive compound iso- 
lated is carboxyl-labeled PGA, followed shortly by 
dihydroxyacetone phosphate (DHAP), fructose 
monophosphates (F.MP) and glucose monophos- 
phate (GMP), both hexoses being 3,4-labeled. Af- 
ter longer exposures of the plant to C'^Oj, radio- 
carbon appears in other carbon atoms of PGA and 
hexose and the distribution of activity is in agree- 
ment with the above conclusions. 



*c + c* 


















Observations on the rate and distribution of la- 
beling of malic acid*"' showed it to be the eventual 
product of a second carboxylation reaction which 
is accelerated during photosynthesis, and it was 
proposed that this second carboxylation played a 
part in the reduction of carbon in photosynthesis, 
leading eventually to the formation of the two-car- 
bon CO2 acceptor (A, above). Malic acid, itself, 
apparently was precluded as an actual intermediate 
by inhibition studies,' but was thought to be an in- 
dicator of an unstable intermediate which was 
actually the first product of the second carboxyla- 
tion. The discovery' of rapidly labeled sedoheptu- 
lose monophosphate (SMP) and ribulose diphos- 
phate (RDP) led to their inclusion in the proposed 
carbon reduction cycle leading to the two-carbon 
CO2 acceptor. 

The reciprocal changes in reservoir sizes of RDP 
and PGA observed when algae were subjected to 
light and dark periods' indicated a close relation- 
ship, perhaps identity, between the RDP and the 
two-carbon CO2 acceptor. 

In order to test these conclusions, it was neces- 
sary to design experiments involving very short ex- 
posures of the plant to C'*02. In some of these ex- 
periments, the C* was administered during "steady 
state" photosynthesis, the environmental condi- 
tions (hght, carbon dioxide pressure, etc.) being 
kept as nearly constant as possible for the hour pre- 
ceding and the time during the experiment. Deg- 
radation methods have been developed for sedohep- 
tulose and ribulose and complete distribution of 
radioactivity within these sugars obtained. 

The results of these experiments seem to obviate 
the possibility that the second carboxylation reac- 

(4) A. A. Benson, S. Kawaguchi, P. M. Hayea and M. Calvin. This 
Journal, 74, 4477 (1952). 

(5) A. A. Benson, et at., "Photosynthesis in Plants," Iowa State 
College Press, Ames, Iowa, 1949, p 381. 

(6) D. W. Racusea and S. Aronoff, Arch. Biochem. Biophys., 42, 25 

(7) J. A. Bassham. A. A. Benson and M. Calvin, J. Biol. Chem., 
18», 781 (1950). 

(8) A. A. Benson, el at . ibid . 196, 703 (1952). 

(9) M. Calvin and Peter Massini, Expcrienlia, 8, 445 (1952). 

tion (leading to malic acid) is a step in carbon reduc- 
tion during photosynthesis. Since no new evi- 
dence has been found for the second "photosyn- 
thetic" carboxylation, it would appear that a carbon 
reduction cycle involving only one carboxylation 
(leading to PGA) is more likely than the previously 
proposed two-carboxylation cycle. 

Experimental Procedure 

Short "Steady State "Eiperiments. — Algae (Scenedesmus 
obliquus) were grown under controlled conditions,' centri- 
fuged from the growth medium, and resuspended in a 1% by 
volume suspension in distilled water This suspension was 
placed in a rectangular, water-jacketed illumination cham- 
ber 6 mm. thick, through which was passed a continuous 
stream of 4% COi-in-air (Fig. 1). From the bottom of the 
chamber, a transparent tube led to a small transparent 
pump constructed of appropriately placed glass valves and 
two 5-cc. glass syringes mounted on a lever arm in such a 
position that the syringe plungers moved in and out recipro- 
cally about 5 mm. when the lever arm was moved back and 
forth by a motor-driven eccentric. The output of the pump 
was divided, the major portion being returned to the illu- 
mination chamber and a smaller portion (20 ml. /minute) 
forced to flow through a length of transparent "Transflex" 
tubing of about 1 mm. diameter and thence into a beaker 
containing boiling methanol. This solvent was found to 
have an apparent killing time of less than 0.2 sec. as deter- 
mined by the cessation of carbon fixation during photosyn- 
thesis. The linear flow rate of algal suspension in the tube 
was about 57 cm. /second. A solution of C'Oj in water 
(0.0716 M, 110 MC./ml.) in a 30-cc. syringe was injected 
through a fine hypodermic needle into the Transflex tubing 
at a point a selected distance from the end of the tubing. 
From the known flow rate of algal suspension in the Trans- 
flex tubing and distance of flow from the point of injection 
of C'*Oj to the killing solution, the time of exposure of the 
algae to C'* was calculated. The flow of the C'Oj-contain- 
ing solution was controlled by driving the syringe plunger 
with a constant speed motor, and the flow rate was 0.5 ml./ 
minute. The resultant dilution of the algal suspension was 
2.5% and the increment in total CO2 concentration less than 

{hot pl>ti r 

Fig. 1. — Schematic diagram of flow system for short exposure 
of algae to C"Oi. 

Since the flow of algal suspension in the tubing was not 
turbulent, some difference in rates of flow at the center and 
at the edge of the tubing was unavoidable. The extent of 
this difference was approximately determined by injecting 
a concentrated dye solution for about 0.5 sec. through the 
hypodermic needle while the flow rate in the tubing was 20 



Bassham, Benson, Kay, Harris, Wilson and Calvin 

Vol. 76 




HC=N— N— CJI. 


t=N— N— C^H, 
I H 



















HC=N— N— C«Hi 

C==N— N— C,Hj + 3HC00H + HCHO 
I H 


1, 2, 3 4. 5, 6 7 



-C 1 

b CHO () 

I i 

•— CH 































ml. /minute and observing the spreading of color during its 
travel through the tubing. For the longest length of tubing 
used, the dye was seen to reach the end of the tubing be- 
tween 14 and 17 seconds, and at a shorter time between 9 
and 11 seconds, so that the spread of flow in time appeared 
to be about 20% of the flow time. The times given are 
average times of exposure of the algae to C*. Use of the 
dye also permitted observation of the mixing of C'Oj solu- 
tkm with algal suspension and mixing time appeared to be 
about 0.2 sec. 

The entire apparatus was illuminated from each side by a 
Bine- tube bank of 40-watt fluorescent lights (white) giving 
a nniform intensity of abwit 2000 footcandles from each side. 
During an experiment the algal suspension was illuminated 
for an hour or more with 4% COj beifort the start of the flow 
C* exposures. Exposures to C"Oj ranging from 1.0 to 16 
sec. were then carried otit and the products of C'^ reduc- 



2, 3, 4, 5. 6 





v + 


tion analyzed in the usual way" by paper chromatography 
and radioautography. 

Short Soybean Experiments. — A single excised trifoUate 
leaf from a soybean plant (var. Hawkeye) was placed in a 
circular flat illumination chamber with a detachable lace. 
The chamber was equipped with two tubes, the lower one 
leading through a stopcock to an aspirator and the "Pper 
one tlmMigh a two-way stopcock to a loop oonUining COj. 
A loosely tied thread led from the leaf stem under the de- 
tachable face gasket, thence through a boiling ethanol bath 
and a glass tube to a weight. The illumination chamber was 
partially evacuated, both stopcocks were closed, and clamps 
removed from the chamber, the detachable face remaining 
in position through atmospheric pressure. With the open- 
ing of the upper stopcock, the C"Oi was swept into the cham- 

(10) A. A. Benson, « •<., This Jootmal, W. 1710 (I960). 


April 5, 1954 

Cyclic Regeneration of Carbon Dioxide Acceptor 


Degradation of Ribulosb 



HC=N— N— CH. 

C=N— N— CHi 
I H 







HC=N— N— CH, 

i=N— N— CH, -f HCOOH 4- HCHO 
I H 


1, 2, 3 4 5 



















ber by atmospheric pressure, the detachable face fell off 
and the leaf was pulled into boiling ethanol. An estimated 
exposure time of 0.4 sec. was obtained. The radioactive 
products were extriicted and analyzed in the usual way. 
In other experiments, longer exposure times were obtained 
by holding the detachable face in position. 

Degradation of Sugars. — The reactions used for the degra- 
dation of the radioactive ribulose and scdohcptulosc are 
shown in the accompany flow sheets 

All radioactive material was purified on two-dimensional 
paper chromatograms.'° Radioactive sedoheptulose was 
converted to the anhydride by liealing at 11)0° with acid- 
treated Dowex-.')0 for one hour, followed by chromatography 
to separate the resulting equilibrium mixture. 

Formation of the Osazones. — The hexosc and hcptose 
osazones were made in the usual manner with phenylhydra- 
zine hydrochloride, sodium acetate and acetic acid. Usu- 
ally about 25 mg. of sugar carrier was used for the reaction. 
Sedoheptulose osazone cocrystallized with glucosazone 
sufficiently well for fructose to be used as carrier with sedo- 
heptulose activity. 

The radioactive arabinosazone was made by the method of 
Haskins, Hann and Hudson" with 10 mg. of arabinose car 
rier. The osazone was recrystallized once and diluted, as 
desired for each degradation, with pure crystalline, non- 
radioactive arabinosazone from a similar large-scale prepa- 
ration . 

Oxidation of Osazones. — The recrystallized osazones were 
treated with periodate in bicarbonate buffer as described by 
Topper and Hastings.'^ The reaction mixture was frac- 
tionated to obtain all the products by centrifuging and 
thoroughly washing the raesoxaldehyde osazone; distilling 
the supernate plus washings to dryness in vacuo and treating 
the distillate with dimedon to obtain the formaldehyde de- 
rivative; and acidifying and vacuum distilling the residue 
to obtain the formic acid, which was counted as barium for- 
mate. All products were recrystallized before counting. 

Cerate Oxidation of Ketoses. — The oxidation of the car- 
bonyl carbon of a ketose to CO2 by cerate ion was performed 
according to the method described by Smith." To a solu- 
tion of an aliquot portion of radioactivity plus weighed 
carrier (sedoheptulosan or fructose) was added a slight ex- 
cess of 0.5 M cerate ion" in 6 iV perchloric acid, the final 
concentration of acid being 4 N. The resultant COj was 

(11) W. T Haskins. R. N. Hano and C. S Hudson. This Jodrnal, 
U, 1766 (1946). 

(12) Y. J. Topper and A B Hastings./. Biol C*«m, 1T9, 1255 (1949). 

(13) G Frederick Smith. "Cerate Oiidimetry." G Frederick Smith 
Chemical Company. Columbua, Ohio, 1942. 

(14) We are indebted to Prof. John C. Speck, Jr , of Michigan State 
College, East Lansing. Michigan, for valuable data and suggestions re- 
garding the use of cerate in these oxidations. 

2HCH0 4- 3HCOOH 
1,5 2.3,4 

swept with nitrogen into COj-free sodium hydroxide. The 
reaction was allowed to proceed for one hour at room tem- 
perature and then the COj was precipitated and counted as 
barium carbonate. In all cases the theoretical amount of 
carbon dioxide was evolved. 

Formation and Oxidation of Sugar Alcohols. — The radio- 
active sugars were hydrogenated with platinum oxide as de- 
scribed previously' and chromatographed on paper for puri- 
fication. Carrier ribitol or voleraitol was added to an ali- 
quot of radioactive alcohol and a slight excess of paraperiodic 
acid was added. The reaction was allowed to stand at room 
temperature for 6-7 hours. Then the formic acid and form- 
aldehyde were distilled off in vacuo. After the formic acid 
was titrated with barium hydroxide, the fonnaldehyde was 
redistilled and precipitated as formyldimedon. Both the 
residue of barium formate and the formyldimedon were re- 
crystallized before plating and counting. 

Bacterial Oxidation of Hepitola from the Reduction of 
Sedoheptulose. — The radioactive reduction products of 
sedoheptulose gave only one spot on chromatography. 
After elution these were oxidized by Acetobacter suboxydans 
in a small-scale modification of the usual method." Two 
mg. of volemitol and about 100 d- of solution of radioactive 
heptitols were placed in a 7-mm. diameter vial. An amount 
of yeast extract sufficient to make a 0.5% solution was 
added. The vial was sterilized, then inoculated from a 24- 
hour culture of Acetobacter and left for a week at room tem- 
perature in a humid atmosphere. 

When the bacteria were centrifuged from the incubation 
mixture and the supernatant solution was chromatographed, 
three radioactive spots were obtained. The two major 
spots were mannoheptulose and sedoheptulose, the oxidation 
products of volemitol. The third had R, values very simi- 
lar to those of fructose and cochromatographed with au- 
thentic guloheptulose'" ( if f in phenol = 0.47; Ri in butanol- 
propionic acid-water = 0.24). After treatment with Do- 
wex-50 in the acid form at 100° for one hour, this third com- 
pound gave a new compound which cochromatographed 
vrith guloheptulosan (Ri in phenol = 0.62; i?i in butanol- 
propionic acid-water = 0.30). It thus appeared that the 
radioactive heptitols are volemitol and 0-sedoheptitol which 
cochromatograph in the solvents used. 

Both mannoheptulose and guloheptulose have carbon 
chains inverted from the original sedoheptulose. In the 
small-scale fermentations, however, the oxidation appeared 
to be incomplete. The original alcohol did not separate 
chromatographically from mannoheptulose. Therefore, 

(16) (a) L. C. Stewart, N. K. Richtmyer and C. S. Hndsoo, TBM 
JoDRNiL, 74, 2206 (1952); (b) we wish to express oor •ppreci»tion to 
Dr. R. Clinton Fuller for his development of the micro-fermentation. 

(16) We wish to thank Dr. N. K. Richtmyer for his generons gift of 
crystalline guloheptulosan. 



Bassham, Benson, Kay, Harris, Wilson and Calvin 

Vol. 76 

the easily purified guloheptulose was used for subsequent 
degradations witli cerate ion, despite its much poorer yield. 

Oxidation of Sedoheptulosan. — The radioactive sarnple 
and carrier were treated with sodium periodate as described 
by Pratt, Richtmyer and Hudson" and allowed to stand at 
room temperature for 3-4 days to give time for most of the 
formate to be released from the intermediate ester. Then 
the mixture was acidified with iodic acid and the formic 
acid was distilled in vacuo. This was then counted as 
barium formate. 


In Fig. 2, the radiocarbon fixed in a "steady 
state" photosynthesis with Scenedesmus is shown as 
a function of time of exposure of the plant to C'^Oj. 

glucose monophosphate and fructose monophos- 
phate ciu-ves although individual points are more 
erratic, probably due to the relative instability of 
the ribulose diphosphate.' The appearance of 
compounds other than PGA with a finite rate of 
labeling at the shortest times is demonstrated in 
Fig. 4 in which the percentage distributions of 
PGA and of the total sugar phosphates are shown. 



e e 10 


Fig. 2. — Radioactivity incorporated in "steady state" photo- 
synthesis with Scenedesmus. 

The rate of incorporation of C'*Os appears to be 
reasonably constant over the period of the experi- 
ment. The distribution of radioactivity among 
various labeled compounds is shown in Fig. 3. The 

a 10 12 


Fig. 3. — Distribution of radioactivity among compounds formed during "steady 
state" photosynthesis with Scenedesmus. 

curve for the sugar diphosphates, principally ribu- 
lose diphosphate, is not shown but lies between the 

(17) J W. Pratt, N. K. Richtmyer aod C. S. Hudson, Tais Joumnal, 
T4, 2200 (ieS2). 

Fig. 4. — Distribution of activity in "steady state" 

The extrapolations of the PGA and sugar phos- 
phates to zero time would give about 75 and 17%, 
respectively. The remaining 8% not shown is dis- 
tributed among malic acid (3%), free glyceric acid 
(2%) and phosphoenolpyruvic acid (3%).' The 
percentage distribution among the sugar phos- 
phates is shown in Fig. 5 where it is seen that no 
single labeled sugar phosphate predominates at 
the shortest times. 

These data alone do not permit 
assignment of an order of preced- 
ence of the various labele(l com- 
pounds in the path of carbon reduc- 
tion. In order to make such an 
assignment it would be necessary to 
measure the relative rates of in- 
crease in specific activity of the 
various compounds. If the slopes 
of the ciu^es shown in Fig. 3 are 
measured between 2 and 10 sec, 
rates of increase in total radioactiv- 
ity are obtained. If these rates are 
divided by the cellular concentra- 
tion of the compounds involved, 
rates of specific activity increase are 
obtained. This has been done using 
measurements of concentrations 
made by two independent'" meth- 
ods which agreed fairly well in rela- 
tive Older {i.e., PGA concentration: 
GMP concentration = 4:1). The 
resulting values ranged from 0.3 for 
GMP to 1.0 for PGA, with FMP, 
DHAP, RDP and SMP falling be- 
tween these values when the rates 
for these compounds were divided by 2, 1, 2, 1, 1 
and 3, respectively, to allow for the number of 
carbon atoms which degradation data reported be- 

(18) A. A. BeiuoD. Z. BUUrochtm. U, 848 (19&2). 


April 5. 1954 

Cyclic Regeneration of Carbon Dioxide Acceptor 


low show to be labeled significantly at these short 
times. This calculation is quite approximate, the 
concentration of compounds involved being meas- 
ured in experiments with algae photosynthesizing 
under somewhat different conditions {i.e., 1% CO2 
instead of 4%). However, such a calculation does 
show more clearly the rapidity with which radio- 
carbon is distributed among the principally labeled 
carbon atoms and the difficulty in assigning an 
order of precedence of labeled compounds on the 
basis of labeling rates alone. 

The fact that compounds besides PGA have fi- 
nite initial labeling slopes (which results in their 
percentage activity not extrapolating to zero at 
zero time) might be explained in several ways. One 
possibihty is that during the killing time some of 
the enzymatic reactions (in this case reduction of 
PGA and rearrangement of the sugars) may not be 
stopped as suddenly as others (the carboxylation to 
give PGA) or may even be accelerated by the ris- 
ing temperature prior to enzyme denaturation. 

Another explanation is that some of the labeled 
molecules may be passed from enzyme to enzyme 
without completely equilibrating with the active 
reservoirs which are actually being measured. This 
sort of enzymatic transfer of radiocarbon could 
invalidate precedence assignments based on rates of 
increase in specific activities since the reservoirs 
would no longer be completely in the line of carbon 
transfer. That the equilibration between reser- 
voirs and enzyme-substrate complexes is rapid com- 
pared to the carbon reduction cycle as a whole is 
indicated by the fact that all the reservoirs become 
appreciably labeled before there is an appreciable 
label in the a- and /3-carbons of PGA, the 1-, 2-, 5- 
and 6-carbons of the hexoses, etc. In any event, it 
would appear to be safer to establish the reaction 
sequences from qualitative differences in labeling 
within molecules (degradation data) and changes in 
reservoir sizes due to controlled changes in one en- 
vironmental variable rather than from quantita- 
tive interpretations of labeling rate data. 

Table I shows the results of degradations on sug- 
ars obtained from the soybean series. The first 
column shows the variation in labeling of carbon 

Table I 
Radioactivitv Distribution in Sugars SEDOHEPTin.osE 






^ .Sedoheptiilose 

C-4 C-1.2.3 C-4.5,G C-7 C-2 

. Hexose 

C-1.7 C-6 C-1,2.3 C-4,.'j.6 










35 12 



28 15 



Fig. 5. — Distribution of radioactivity incorporated in 
"steady state" photosynthesis with Scenedesmus: ©, sedo- 
heptulose phosphate; 9, glucose phosphate; ®, dihydroxy- 
acetone phosphate; O, fructose phosphate. 

since the carbon dioxide is depleted just prior to the 
administration of C'Oa. Included in the table is 
a complete degradation of a sedoheptulose sample 
from Sedum speclabile grown in radioactive carbon 
dioxide for two days (kindly supplied by N. E. 
Tolbert, Oak Ridge National Laboratory). As- 
suming this sample is uniformly labeled, its degra- 
dation indicates the probable limits of accuracy of 
the other degradations — about ± 10% of the ob- 
tained value, mainly due to plating and counting 
errors resulting from the low amount of radioactiv- 
ity available for degradation. The five degrada- 
tions on sedoheptulose make it possible to obtain 
separate values for all the carbon atoms. Although 
the carbon-fourteen labels of carbon atoms 1 and 
were not determined in the case of the Scenedesmus 
experiments, they were assumed small and approxi- 
mated equal to carbon-fourteen labels found in 
carbons 2 and 7, by analogy with the soybean leaf 
experiments where the labels of all carbon atoms of 
the sedoheptulose were determined. The label in 
each carbon atom of the ribulose can be obtained 
individually from the three degradations performed. 
The distributions in Table II should be interpreted 
as a clear qualitative picture of the position of the 
radioactivity within the molecule rather than as a 

Table II 

Radioactivity Distribution in Compounds from Flow 
Experiments (Algae) 





-5.4 Seconds- 

















8 5 Seconds 

Sedohep- Ribu- 

tulose lose 







number four of sedoheptulose obtained from soy- 
bean leaves exposed to C'''02 for very short periods. 
These soybean leaf experiments are, of course, not 
intended to represent "steady state" photosynthesis 

quantitative picture. Fewer points were taken in 
this "steady state" flow experiment than in the 
one described earlier in order to obtain more la- 
beled sugar per point for degradation purposes. 



Bassham, Benson, Kay, Harris, Wilson and Calvin 

Vol. 76 

In other experiments" the Scenedesmus have been 
kept at a steady state of light, temperature, CO2 
pressure, etc., and constant C'K)2 specific activity 
until successive samplings of the suspensions showed 
uniform labeUng ("satiu-ation") of all the common 
photosynthetic reservoirs (PGA, RDP, GMP, etc.). 
The total CO2 pressure was then rapidly changed 
from 1% C02-in-air to 0.003% in air, all other en- 
vironmental conditions, including the specific ac- 
tivity of C"02, being kept constant. The condi- 
tions of this experiment were, therefore, similar to 
those used previously' to study changing steady 
state except that CO2 pressure was changed in- 
stead of illumination. In the case where the CO2 
pressure was lowered (Fig. 6), the initial effects on 
the reservoir sizes of PGA and RDP were just the 
opposite of those observed when the illumination 
was stopped. Lowered COi pressure resulted in an 


Triose phosphate 


B -«- 

A -*- 





1% CO, 







45 minutes C"0, at 6° C. 
Fig. 6. 


Time in seconds. 

increase in the reservoir size of RDP and a decrease 
in that of the PGA. After a time the reservoir of 
RDP passed through a maximum and dropped to 
a lower level but the new steady state RDP res- 
ervoir was now greater relative to that of PGA. 
The labeled glycolic acid present, though rather a 
small percentage of total activity, increased many 
fold when the COa pressure was lowered. The res- 
ervoir of glycolic acid increased much more slowly 
than that of the RDP and did not pass through a 
corresponding maximum, thus eliminating the pos- 
sibility that most of the labeled glycolic acid was 
formed by thermal decomposition of RDP subse- 
quent to killing of the cells. 

1. Origin of PGA. — It has been suggested that 
RDP is the compound which supplies the two- 
carbon atoms for the carboxylation reaction lead- 
ing to PGA.' If the reactions of these compounds 
are represented by 

(19) A. T. Wilioo, Thesis, to be submitted as partial fulfillment ol 
requirements for the degree of Doctor of Philosophy. UnlTenltjr of 

Sugar rearrangements 

then the initial changes in reservoir sizes which 
would accompany changes in light or COj pressure 
can be predicted. When the light is tmned off, 
reducing power [H] decreases, so the reservoir of 
PGA would increase and that of RDP decrease. If 
CO2 pressure decreases, then the reservoir of RDP 
would increase and that of PGA would decrease. 
Both effects, as well as those opposite effects which 
would be expected to accompany a resumption of 
light or increase in COj pressure, 
have been observed. These re- 
sults support the proposal of a 
carboxylation of RDP to give 
two molecules of PGA or the 
reductive carboxylation to give 
one molecule of PGA and one 
of phosphoglyceraldehyde as 
the first step in the path of 
carbon dioxide reduction. 

It is also possible that the 
products of this carboxylation 
may be phosphoglyceraldehyde 
and 3-phosphohydroxypyru- 
vate. In this case subsequent 
reduction of the phosphohy- 
droxypyruvate would give first 
PGA and then phosphoglycer- 
aldehyde. The reaction of phos- 
phoglyceraldehyde with hy- 
droxypyruvate to give ribulose 
monophosphate and COj has 
been demonstrated by Racker^° 
to take place under the influ- 
ence of the transketolase en- 
zyme. However, the increase in PGA concentra- 
tion which is observed on stopping the illumina- 
tion of photosynthesizing algae,' would probably 
not be seen if a reduction of hydroxypyruvate 
were required to form PGA since the reducing agent 
would presumably no longer be formed in the dark. 
Moreover, paper chromatographic analysis should 
detect either phosphohydroxypyruvate or its de- 
carboxylation product, phosphoglycolaldehyde, and 
neither have been found in our experiments. When 
C'*-labeled hydroxypyruvate was administered to 
algae in this Laboratory, the labeled acid was me- 
tabolized to give a variety of compounds, similar to 
those formed from labeled pyruvate or acetate, which 
were related more closely to the tricarboxylic acid 
cycle and fat synthesis than to the compounds usu- 
ally associated with carbon reduction in photosyn- 

There remains the possibility that the RDP first 
spUts to give a three-carbon molecule and a free 
two-carbon fragment which is then carboxylated. 

(20) B. Racker, G. de la Haba and I. G. Leder, This Joijknal, Ti. 
lOlO (1068). 

0003% COi 




April 5, 1954 

Cyclic Regeneration of Carbon Dioxide Acceptor 


However, if the glycolic acid is an indication of the 
free two-carbon fragment, then the observation 
that its increase in concentration (following reduc- 
tion in COj pressure) is not as rapid as the increase 
in RDP concentration suggests that the Cj com- 
fjound is not as closely related to the carboxylation 
reaction as the RDP. 

2. Origin of Ribulose Diphosphate. — If one 
considers the principal labeling at short times of 
PGA,^ RDP, SMP and the two hexose monophos- 
phates^ as, respectively 























it apj>ears that the ribulose is not derived entirely 
from a Ce ^- Ci -f- C6 split or a C7 — >• C2 + Ce split. 
No five carbon fragment of the hexose or the hep- 
tose molecules contains the same distribution of 
radiocarbon as ribulose. The combination of C3 
with a labeled C2 fragment could account for the 
observed radioactivity. However, some mecha- 
nism for the labeling of the C2 fragment would be re- 
quired. One such mechanism would be the break- 
down of hexose simultaneously into three Cj frag- 
ments,^' and since carbon atoms 3 and 4 of hex- 
ose are labeled, a labeled Cj fragment might thus 
be obtained. To our knowledge there exists no 
precedent as yet for this type of reaction. 

Another way of accounting for the observed dis- 
tribution of radioactivity which seems quite plaus- 
ible in view of the rapidly accumulating enzymatic 
evidence for the reverse reaction '''•^'"-■' is the forma- 
tion of ribulose from sedoheptulose and triose. This 
reaction could result in the observed labeling 


I I 

=0 + CHOH 







CHjO© phospho- 
SMP glyceraldehyde 



c— o -t- 










CH2O© J 










monophos- monophos- 
phate phate 

If the ribose-5-phosphate and ribulose-5-phosphate 
are then converted to RDP the resulting distribu- 

(21) H. GaffroD, E. W. Fager and J. L. Rosenberg, "Carbon Dioxide 
Fixation and Photosynthesis," Symposia of the Society for Experi- 
mental Biology (Great Britain), Vol. V, Cambridge University Press, 

(22) B. Aadrod, R. S. Baudurslii, C. M. Greiner and R. Jang. J. 
Biol. Chem.. SOI, 619 (1953). 

(23) B. L. Horecker and P. Z. Smymiotis, This Journal, T4, 212S 

(24) B L Horecker and P Z. Smymiotis, itruf , It, 1009 (1963). 

tion of label would be that observed (carbon skele- 
ton at right of reaction). 

3. Origin of Sedoheptulose.— The degradation 
data appear to eliminate the possibility of formation 
of sedoheptulose by a simple 6 + 1 or 6 -f 2 addi- 
tion, if we assume that no special reservoirs of pen- 
tose and hexose exist with distributions of radioac- 
tivity different from those measured. A reverse 
of the reactions proposed above for formation of 
RDP would require segregation of ribose and ribu- 
lose distributions as well as some other mechanism 
for labeling the ribose in the manner shown. It 
does seem likely that all the reactions involving 
rearrangements of sugars and perhaps those in- 
volving reduction of PGA as well are at least par- 
tially reversible in the time of these experiments. 
If all these compounds are intermediates in a cycle 
of carbon reduction, then during steady state pho- 
tosynthesis there will be a net "flow" of radiocarbon 
in the "forward" direction, but the possibility that 
the distribution of radiocarbon in later intermedi- 
ates may reflect to some extent that of earlier inter- 
mediates cannot be entirely ignored. 

The condensation of a triose with a C4 fragment 
would give the observed distribution if the C4 frag- 
ment is labeled in the carbon atoms 1 and 2 


















Enzymatic evidence for this reaction and its re- 
verse has been reported. ^''^^ 

4. Origin of the Four-Carbon Fragm«it. — Two 
possible modes of formation of the four-carbon 
fragment with the above labeling are a Cj + C3 
addition, and a Ce -+ [C2] -|- [C4] split. The C, + 
C3 addition which leads to malic acid produces a 
C4 fragment labeled in the two terminal positions." 
Therefore, the reduction of the dicarboxylic acid 
formed as a precursor to malic acid could not result 
in a C4 fragment with the C'^ distribution required 
for the formation of 3,4,.5-C'* labeled sedoheptulose. 
The rapid introduction of radiocarbon into malic 
acid in earlier experiments* can be accounted for if 
it is assumed that the reservoir size of malic acid, 
depleted during the air flushing prior to the addi- 
tion of HC'HDa", was increasing after the addition 
of radiocarbon due to the increase in total CO2 
pressure. Also, after the carboxyl group of PGA 
and phosphoenolpyruvic acid have become appre- 
ciably labeled, the mahc acid is doubly labeled. 

It is interesting to note that in the long term 
"steady state" experiments in which the light was 
turned off,' the mahc acid concentration dropped 
when the light was turned off rather than increas- 
ing as PGA concentration increased. If maUc acid 
were an indicator of a four-carbon intermediate in 
carbon reduction, the product of a second carboxyl- 

(25) B L. Horecker and P Z Smymiotis, ifriJ, 76, 2021 (1853). 



Bassham, Benson, Kay, Harris, Wilson and Calvin 

Vol. 76 

ation, then one would expect its concentration to 
increase in the dark for two reasons. First, there 
no longer is reducing power which would reduce 
the carboxylation product to sugar if this product 
were an intermediate in CO2 reduction. Second, 
the rate of formation of malic acid should increase 
since this rate depends on the CO2 concentration 
(which remains constant), and the concentration 
of phosphoenolpyruvic acid (which increases paral- 
leUng the PGA concentration). The decrease in 
malic acid concentration could be easily explained 
on the basis of the proposed light inhibition of py- 
ruvic acid oxidation.' The cessation of illumina- 
tion should permit increased pyruvic acid oxidation, 
thus providing more acetyl-CoA, which can react 
with oxaloacetic acid derived from malic acid. 

It is possible that there is a different "second 
carboxylation" (Cj + Ci) leading eventually to a 
four-carbon fragment which can react with those 
to give sedoheptulose, but there seems to be no 
evidence whatever for such a reaction at present. 
Moreover, such a reaction should lead in short 
times to a four-carbon fragment somewhat more 
labeled in the terminal carbon position than in the 
second carbon position due to dilution of the carbon 
introduced in the first carboxylation reaction by 
the PGA and triose reservoirs. This is not the 
case — in fact in the very shortest times the ter- 
minal carbon position of the hypothetical d frag- 
ment (carbon four of sedoheptulose) is actually less 
labeled than the second position, at least in the soy- 
bean experiments. 

The most likely source of the C4 fragment seems 
to be a Co -► [C*] -f [C2] split. Trioses could then 
react with [C4] and [C2] to give sedoheptulose and 
ribulose, respectively. One possible formulation of 
these reactions would be 




=0 -I- CHOH 




+ c=o — 



c=o -f 














The first reaction as written above would be a 
transketolase reaction of the type reported by 
Racker, et al.,^" who found that this enzyme splits 
ribulose-.")-phosphate, leaving glyceraldehyde-.3- 
phosphate and transferring the remaining two 
carbon atoms to an acceptor aldehyde phosphate of 
2-, 3- or 5-carbon atoms. No mention was made of 
the effect of transketolase on ribulose-5-phosphate 
with erythrosc-4-phosphate which would result in 

the formation of fructose phosphate by a reaction 
which is just the reverse of the Ce-*- [C2] + [Ci] split 
written above. ^' 

The labeling of carbon number 4 in sedoheptulose 
observed in the case of the very short periods of 
photosynthesis with soybean leaves seems to cast 
some doubt on the Cj -»■ [C2] -f [C4] split unless 
one can assume that the Ce which splits is itself 
not symmetrically labeled at the shortest times, due 
to different specific activities of the two trioses 
which react to give hexose 







, I 

- incomplete - 




later, hence 
more complete 




. I 

" I 




















CHjO© I 


c=o C=0 



••CHOH < 



Degradation of fructose from the 0.4- and 0.3- 
sec. experiments showed no significant difference 
between the two halves of fructose. It is quite 
possible, however, that the differences in denatura- 
tion rates of various enzymes mentioned earlier 
may influence the results in these short times. 

Combining these reactions with others aheady 
proposed we have the following cyclic path of car- 
bon reduction during photosynthesis. The car- 
bon fragments specified only by the number of car- 
bon atoms in their chains are all at the sugar level 
of reduction 

3Ci -I- 3C0j — 


2C, - 

C, -I- 2C, - 

C, + C, 


> 6C, 


Cs-h C, 

The net reaction for each turn of the cycle is 

12 (HI -f 3C0,- 

C.H,0, -I- 3H,0 

The operation of this cycle is illustrated in Fig. 7. 

5. Energetics of the Carbon Reduction Cycle. — 
That the enzymatic rearrangements of sugars re- 
quires no additional supply of energy in the form 
of ATP or other sources seems to be indicated by 
the experiments with isolated and partially purified 
enzyme preparations in which such rearrangements 
have been carried out without the addition of 
energy donors. The free energy change of the car- 
boxylation reaction can be roughly estimated. Es- 
timating the free energy difference between ribose- 

(26) Since this was written, a private communiration from Dr. 
Racker has informed us that he has observed this reaction with F-6-P 


Cyclic Regeneration of Carbon Dioxide Acceptor 
Chl* [0] 





Fig. 7. — Proposed cycle for carbon reduction in photosynthesis. Heavy lines indicate transformations of carbon com- 
pounds, light lines the path of conversion of radiant energy to chemical energy and the subsequent use of this energy 
stored momentarily in some compound (E), to form a reducing agent [H] and oxygen from water. 

.5-phosphate and RDP equal to that between GMP 
and fructose diphosphate, the free energy change 
for the reaction below is about —7 kcal.^'-^ 




CHOH + CO, + H,0 




2CHOH + 2H-' 



(5 X \()-' M) (10-«A/) 

(1.4X10-»Af) (10-' W) 

Af 2© -176 -95 

-57— 2(© -158) 2(-95) 

AF - -7 kcal. 

In the above calculation the concentrations of 
RDP and PGA measured with Scenedesmus during 
photosynthesis with 1% COz' are used. The mech- 
anism of the reaction may consist of the addition of 
CO2 to the 2,3-enediol sugar formed by enolization 
of the RDP. The intermediate compound would 
be 2-carboxypentulose-3. The free energy for the 
formation of the ion of this acid and H+ {pH 7) 
from COi and RDP is estimated as zero when the 
concentration of the intermediate acid is 10~' M. 
Subsequent hydrolytic splitting of this compound 
to two molecules of PGA and another hydrogen ion 
would proceed with a free energy change of —7 

The energy required to maintain the operation of 
the proposed carbon reduction cycle might be sup- 
plied entirely in the reduction of PGA to triose 
phosphate. If this reduction were accomplished 
by a reversal of the enzymatic reaction usually writ- 

(27) The intemal energy of the -PCHH" group, exclusive of the 
energy of bondiog to the remainder of the molecule ia here denoted by 
© and assumed constant throughout. 

(28) J. A. Bassham, Thesis, submitted as partial fulfillment of re- 
quirements for the degree of Doctor of Philosophy, University of Cali- 
fornia, 194B. 

ten, each "turn" of the cycle would be represented 
by three times the reaction 

2DPN[H,1 -I- 2ATP -|- CO, — *■ |CH,0| -f (A) 
+ 2DPN -t- 2ADP -t- 2© -f H,0 

This is the sum of the reactions 

2[DPN1H,1 + V1O. — »- 

DPN -I- H,0) AF - -101 lead. (B) 

2[ATP — »-ADP-f©I A^ - -21 kcal. (C) 

CO, -f H,0 — *- O, -f- |CH,0| HF " -1-116 kcal. (D) 

The efficiency of the transfer of energy of reactions 
B and C to reaction D is 116/(21 + 101) = 0.96. 

However, additional energy might be supplied 
to the operation of the cycle by phosphorylation 
reactions in which additional molecules of ATP are 
required. One such reaction may well be the phos- 
phorylation of ribulose monophosphate to give ribu- 
lose diphosphate. In this case, one additional 
molecule of ATP would be required per molecule 
of CO, reduced. The efficiency of the net reaction 
(A') would then be 116/132.5 = 0.88. 

2DPN[H,) -I- 3ATP + CO, >- 

|CH,0| -I- 2DPN -f- 3ADP -f 3© -t- H,0 (A') 

The over-all efficiency of photosynthesis would 
be the product of 0.96 or 0.88 and the efficiency of 
the process by which water is photolyzed to give 
oxygen with the production of reducing power, fol- 
lowed by the conversion of the energy of this re- 
ducing power to DPN[H,] and ATP. 

If the mechanism for photolysis of water in- 
volves thioctic acid, as has been proposed,^ the 
energetics of the photochemical and following steps 
can be estimated 

[Y + HOH -^ If 


S — S 


(29J J. A. Barltrop, P. M. Hayes and M. Calvin, to be pnblbhed. 



Bassham, Benson, Kay, Harris, Wilson and Calvin 

Vol. 76 

(where the symbol / represents the side chain: 
— (CH2)4COjH). 

+ 1 I + HjO + 'AO, (F) 

In this process, two quanta are required for each 
dithiol molecule formed. The stored energy is the 
sum of the energies of the two half reactions 

2H+ + 2e- + "AOj Af = +37.5 kcal. (G) 



+ 2H 
S — S 
which is 

H,0 + 

+ 2e-- 


E =• -0.3 v.» 
AF= +13.8 kcal. 

SH (H) 

S — S 

(2A») {\^ 

- — ► I I +'AO, 



51.3 kcal. 


Since the energy available from two light quanta at 
7000 A. is 2 X 40.7 or 81.4 kcal., the efficiency of 
this process would be 51.3/81.4 = 0.63. 

If Co-I is used in the reduction of PGA, the re- 
duced coenzyme could be formed with high efficiency 
from the dithiol 


DPNlH.l + S — S 

LP = -0.8 kcal. (J) 

The required ATP could be formed in some way 

by oxidation of SH SH or DPN [Hi] by an ener- 
getic coupling of the reactions 

DPNIH,] + 'AO, — *■ DPN + HjO 

AF = -50.5 kcal. (K) 
ADP + © > ATP Af - +10.5 kcal. (L) 

Since from one to four molecules of ATP might be 
formed per DPNfHj] oxidized, a wide range of ef- 
ficiencies would be possible. A value of three has 
been suggested*' and if this is used, the resulting 
coupling reaction could be written 

DPN[H,] + VtO» + 3ADP + 3© *- 

DPN + H,0 + 3ATP (M) 

Multiplying reaction J by 3 and combining with 
reaction M we have 


3SH SH + 2DPN + 3ADP + 3© + 'AO, — *■ 

(SO) I. C. GunsaJus, Bymposlnm oa "Mechaaism of Bazyme Ac- 
tion," McCollum-Pratt Inatitute. Johiu Hopkins Univcnity, 1033, to 
be published. 

(31) A, L. Lehnioger, "Phosphorus Metabolism," Vol.1, Johns Hop- 
kins University Press, 1961, page 344. 

+ 2DPN[H,] + 3ATP 

+ H,0 + 3S — S 


in which the stored energy is 132.5 kcal. and the en- 
ergry expended is three times reaction I = 154 kcal. 
The efficiency of the energy transfers represented 
by reaction N is then 132.5/154 = 0.86. 

Combining the efficiencies of reactions A', I 
and N results in a calcidated over-all efficiency for 
photosynthesis of 0.88 X 0.63 X 0.86 = 0.48. 
Since Uie mechanism outlined above would require 
six quanta for each molecule of carbon dioxide re- 
duced (two quanta for each molecule of dithiol used 
in reaction N) this efficiency can be obtained di- 
rectly from the energy of these quata (244 kcal.) and 
the energy of reaction D: 116/244 = 0.48. 

Higher apparent efficiencies would be obtained 
at low light intensities where the dark internal con- 
version of prior storage products (involving no net 
uptake of oxygen or evolution of COj) would sup- 
ply appreciable amounts of ATP, DPNH, reduced 
thioctic acid and possibly intermediates of the Oj 
evolution chain as well." 

Since reaction I as written stores only 51.3 kcal. 
of 81.4 kcal. available, it is posable that some 
mechanism may exist for the storage of some of this 
energy in the form of either additional reducing 
power or high energy phosphate. In this case, the 
over-all efficiency would be higher. 

6. Other Biological Evidence. — The intercon- 
versions of the five-, six- and seven-carbon sugars 
are being investigated by several laboratories. The 
postulated cychc reactions which our data suggest 
are consistent with the observations of these various 
groups. Both the work of Axelrod, et al.,'* with 
spinach preparations and the results reported by 
Dische and Pollaczek" with hemolysates demon- 
strate the sequence 

ribose phosphate — >■ heptulose phosphate + 

triose phosphate — *■ hexose phosphate 

Recently studies have been made of the distribu- 
tion of C" in products resulting from conversion of 
l-C* labeled pentoses. Neish" has studied the 
products of bacterial metabolism of several pentoses 
while Wolin, et al.,** investigated the products of 
enzymatic conversion of ribose-5-phosphate. In 
both cases, the distribution of radioactivity in the 
products coidd be accounted for by a reversal of 
the reactions herein suggested, although a limited 
number of other interpretations of their data are 


(32) Z. Dische and B. Pollaczek, paper presented at Second Inter- 
national Congress of Biochemistry, Paris, France. 1952. 

(33) A. C. Neish. paper presented at American Society of Bacteriolo- 
guts Meeting. San Francisco, Calif., 1953. 

(34) H. B. Wolin. B. L. Horecker, M. Gibbs and H. Klenow, paper 
presented at Meeting of American Institute of Biological Sciences, 
Madison, Wisconsin, 1963. 





Lawrence Radiation Laboratory, University of California, Berkeley, Calif. {U.S.A.) 

(Received January 30th, i960) 


Kinetic studies have been made of the rates of appearance of ^*C in individual com- 
pounds formed by Chlorella pyrenoidosa during steady state photosynthesis with 
"COjj. These rates have been compared with rates of COj and ^*C disappearance from 
the gas phase during the same experiments. 
The following results were obtained : 

1. After the first few seconds, the rate of appearance of ^*C in compounds stable 
to drying on planchets at room temperature is 95 to 100 % of the rate of uptake of 
carbon from the gas phase. 

2. After the first few seconds, the rate of appearance of carbon in compounds 
isolable by usual methods of paper chromatography constitutes at least 73 to 88 % 
of the rate of uptake of carbon from the gas phase. Compounds formed from the 
carbon reduction cycle via the carboxylation of ribulose diphosphate account for a 
least 70 to 85 % of the uptake, while carboxylation of phosphoenolpyruvic acid 
appears to account for at least another 3 %. 

3. The induction period in the appearance of ^*C in stable compounds may be 
due to a reservoir of intracellular COj and HCO3 or to some other volatile or unstable 
compound. If so, this reservoir contains no more than 1.5 )umoles of carbon, corre- 
sponding to about 7 sec carbon fixation in the experiment in which it was measured. 

4. No other carboxylation reactions, such as the carboxylation of y-aminobutyric 
acid, could be observed. The rate of labeling of glutamic acid after 5 min of exposure 
of the algae to i*CO, reached a maximum rate of about 5 % of the total uptake rate, 
but this labehng appears to be due to conversion of labeled intermediates formed 
from the carbon reduction cycle or phosphoenolpyruvic acid carboxylation. 

5. The in vivo carboxylation of ribulose diphosphate in the light appears to be 
followed by conversion of the product to one molecule of phosphoglyceric acid, 
containing the newly incorporated ^^COj and one molecule of some other (kinetically 
distinguishable) three carbon compound. This reaction would be different from the 
one reported for the isolated enzyme system and the in vivo reaction in the dark, 
which produces two molecules of 3-phosphoglyceric acid. 

Abbreviations: PGA or 3-PGA, 3-phosphoglyceric acid; PEPA, phosphoenolpyruvic acid; 
RuDP, ribulose 1,5-diphosphate; ATP, adenosine triphosphate; TPNH, reduced triphospho- 
pyridine nucleotide. 




Much of the biochemical pathway through which carbon dioxide is reduced during 
photosynthesis in algae has been established^-^ A principal feature of this pathway 
is the carbon reduction cycle. A simplified version of this cycle is given in Fig. i, 
which shows the key steps. 

To map these paths, Calvin et al.^^" gave radioactive compounds, such as 
"CO2 and KHj^^po^, to photosynthesizing plants. The plants made various reduced 
organic compounds from these labeled substrates. They were then killed and the 
soluble compounds were extracted from the plant material and analyzed by two- 
dimensional paper chromatography and radioautography. The compounds were 
identified and their radioactive content determined. From the amount and location 
of radioactive elements within compounds following exposures of the plants for 
various lengths of time and under various environmental conditions, biochemical 
pathways were followed. 

Fig. I. Carbon reduction cycle (simplified version), (i) Ribulose diphosphate reacts with COj to 
01-.OL1GO-. AND GLYCEROL PHOSPHATES givc an unstable six carbon compound which 

poLrssccHABiDES GALACTOSE PHOSPHATES spUts to give two three carbon compounds. At 

least one of these is 3-phosphoglyceric acid. The 
other three carbon compounds might be either 
' ^'^^ 3-PGA, as it is known to be in the isolated en- 

PENTOSE-5-PHospHATES fHEPTOSE PHOSPHATES zyme System, or some other three carbon com- 

4TP (4 <HExosE PHOSPHATES pound such as a triose phosphate (dashed arrow) . 

\| JTRiosE PHOSPHATES ^^j pQ^ jg reduced to triose phosphate with 

RIBULOSE DIPHOSPHATE _^ ^^/ / j^jp ^^^ TPNH derived from the light reaction 

.^^TPNH 2^OTP^ ^^^ water. (3) Various condensations and re- 

arrangements convert the triose phosphates to 
pentose phosphates. (4) Pentose phosphate is 

c .„„„j£ phosphorylated with.\TP to give ribulose di- 

-ALANINE phosphate. Further carbon reduction occurs via 

conversion of PGA to phosphoenolpyruvic acid, 
(s) andcarboxylation, (61, to form a four carbon 
compound (probably oxaloacetic acid). Keac- 
-ASPARTicAcio tions leading to the formations of some of the 

secondary intermediates in carbon reduction are shown by the arrows lettered a through g. 

In the present study we have extended our information about these pathways 
by more precise control of the environmental conditions during exposure of the 
plants to tracers. At the same time we have made measurements of the rate of entry 
of tracer into the plant and of the rate of appearance of the tracer in specific com- 

We sought answers to the following questions: (a) How much of the total carbon 
taken up by the plants enters the metabolic network via carboxylation of ribulose 
diphosphate (reaction i)? (b) How much of the total carbon taken up enters by 
carboxylation of PEPA (reaction 6)? (c) Are any other carboxylation reactions, such 
as the carboxylation of y-aminobutyric acid", of any importance in steady state 
photosynthesis? (d) Does the carboxylation of ribulose diphosphate in vivo lead to 
one product only (PGA) or does it lead to two products (PGA and some other 3-carbon 

"Steady state photosynthesis" as used in this paper, is defined as a condition 
under which unicellular algae are carrying out the reaction of photosynthesis, are 
synthesizing all of the normal cell constituents, and are growing and dividing at 

aTP =■ 




constant rates during the course of the experiment. Moreover, the rates of photo- 
synthesis in experiments which will be reported here were between 30 and 80 % of 
the maximum rates at which these algae are capable of photosynthesizing at room 


Plant material 

The plants used in all experiments were the unicellular green algae, Chlorella 
pyrenoidosa, raised in continuous automatic culture tubes as described previously*. 
The algae were raised and harvested as a 0.5 °o (volume wet packed cells/volume) 
suspension. The algae were centrifuged from the culture medium and then suspended 
in a special nutrient solution (described later). This suspension (80 ml) was placed 
in the illumination chamber of the steady state apparatus. 

Fig. 2. Steady state apparatus. (1) algae chamber. (2) water or nutrient solution reservoii, (3) acid 
or base reservoir, (4) pH electrodes, (5) solenoid operated pH control valve, (6) solenoid operated 
sampling valve, (7) small lamp, (8) photovoltaic cell, (9) large gas reservoir, (10) four-way stopcock. 

Steady state apparatus 

In the steady state apparatus, shown schematically in Fig. 2, a stream of gas 
(i to 2 % CO 2 in air) is cycled through a closed system. The gas is bubbled through 
the 0.5 % or 1.0% suspension of algae (80 ml) at a rate of approximately 1 1/min. 
Gas and liquid mix rapidly in the algae chamber, which is 3/8" thick and 4" in 
diameter (inside dimensions). The algae chamber is illuminated from both sides 
by G.E. RSP2 photospot incandescent lights through an infrared absorbing glass in 
a water bath, or in some experiments from one side by an incandescent lamp and 
from the other side by a bank of eight 8", 6 W fluorescent lamps (blue and cool white). 
In either case, the voltage to the incandescent lamps is adjusted just to give hght 
saturation of the oxygen evolution rate. The algae chamber is water jacketed, and 



the water is circulated in a thermostated bath. The temperature of this bath is set so 
that during steady state photosynthesis the temperature indicated by the thermo- 
meter in the algae suspension reads 25°. 

The algae chamber is connected to a side loop through which the algae suspension 
is made to circulate by the flow of gas into the chamber. A beam from a small lamp 
passes through a window in the side loop to a photovoltaic cell which measures the 
light absorption and hence the density of the algae. Electrodes in the side loop 
measure pH, which is recorded on a multipoint recorder. The pH meter output is also 
connected to a control relay which, through the activation of a solenoid-operated 
valve, can cause acid or base from a reservoir to be added in small volumes to the 
algal suspension. Another reservoir within the closed system contains distilled water 
or nutrient solution, which can be added to the algal suspension to dilute it to the 
selected concentration as the algae grow. 

A solenoid-operated sampling valve at the bottom of the chamber permits one 
to take I -ml samples rapidly (every 2 sec if desired). The inside of the algal chamber is 
maintained at slightly above atmospheric pressure to force the algal sample out of 
the chamber. When samples of algae are taken, they are run into 4 ml of methanol 
at room temperature. This gives a mixture which contains about 80 % methanol 
by volume. No significant difference in the resulting labehng pattern is seen whether 
the algae are killed this way, in boiling ethanol, or in ethanol kept at — 40°. 

After the gas in the closed system bubbles through the algae, it passes through 
instruments which measure COg, ^*C, and Og, and each measurement is automatically 
recorded. From the known sensitivities of these instruments and the volume of the 
system, one can calculate rates of exchange of these quantities and specific radio- 
activity. A large reservoir and small reservoirs may be connected or disconnected 
from the closed system to obtain closed systems of various sizes. The volume of 
the largest system is 6400 ml, while the volume of the smallest system is 435 ml. The 
system can be open during the pre-labeling period by means of a stopcock. 

Nutrient solution 

For steady state experiments it is necessary to supply the algae with all the in- 
organic compounds required for them to photosynthesize and grow at a normal rate. 
Unfortunately, the nutrient solution in which they are usually grown in the laboratory 
contains quantities of salts which make impossible an adequate separation of labeled 
compounds by two-dimensional paper chromatography. Therefore, the algae are 
suspended in much more dilute nutrient solutions of which that in Table I is typical. 



(NH^JijHPO^ 40 mg/1 

MgSOi-yHjO 2omg/l 

NH4CI 20 mg/1 

KNO3 20 mg/1 

Arnon's A-4 solution of trace elements plus 

CoClj-6HjO (40 mg/1) and M0O3 (15 mg/I)i2 i ml/1 

Fe'''+-versenol solution to give 90 xaM Fe++ i ml/1 

NH.VOj (23 mg/1) I ml/I 



This medium was adequate to maintain nearly a constant rate of photos3aithesis 
in experiment steady state No. i8. In other experiments, such as steady state 28, the 
algae growing under steady state conditions would in time exhaust the supply of 
ammonium ion contained in this medium. However, it has been observed that as the 
algae take up ammonium ion, the pH of the medium tends to decrease, presumably 
due to the exchange into the medium of hydrogen ions for ammonium ions. Therefore, 
dilute NH4OH was added to the algae suspension automatically by the pH control 
system, thereby maintaining constant pH. At the same time ammonium ion concen- 
tration was maintained approximately constant. The nutrient solution for pH 
control was diluted by trial and error until its addition kept the algae density constant. 
To it were added other inorganic ions in a ratio to the ammonium ion which was 
estimated to provide the algae with an adequate level of these ions for growth for a 
limited period. The resulting pH control medium used in steady state experiment 28 
is shown in Table II. 



(NH^jHPO^ 6.6 mg/1 

(NHi)jS04 6.6 mg/1 

NH4OH 0.55 mg/1 

FeClj-6H20 50 mg/l 

KCl 8.0 mg/1 

Trace elements as in starting medium 

Administration of^*C 

During the first part of the experiment the algae are kept photosynthesizing in 
the light with a constant supply of 1.5 to 2 % unlabeled CO^ in air for 0.5-1 h. Constant 
pH, temperature, and light intensity are maintained during this time, and during 
the subsequent exposure to "COg. In the experiments reported here the pH was kept 
at 6. Rate measurements of CO2 uptake and O2 evolution are made by making the 
closed system small, 435 ml for a few minutes, and observing the rate of change of 
CO2 and O2 tensions as indicated on the recorder. The closed gas system is made 
large again, and at zero time, ^^COg is added to the system by turning a stopcock. 
At the same instant a solution of NaH'^COj is injected directly into the algal suspen- 
sion. The amount and specific radioactivity of the injected bicarbonate solution is so 
calculated that it will immediately bring the specific radioactivity of the dissolved 
COg and bicarbonate already present in the algal suspension to its final value. This 
is the specific radioactivity which will obtain for all the CO 2 and bicarbonate in the 
gas and liquid phases of the closed system after complete equilibration has occurred. 
An example of this calculation is given in Table III. Samples of the algae suspension 
of uniform size are taken every 5 or 10 sec for the first few minutes, and then less 
frequently for periods up to i h. Each sample is taken directly into 4.0 ml of methanol 
(room temperature) in a centrifuge tube (preweighed) . Sample tubes are reweighed 
to give the sample size (± i %). After an hour at room temperature, the samples are 
centrifuged and the 80 % methanol extract removed, i ml of methanol is added to 
the residue and stirred a few minutes, then 4 ml water is added and the mixture 










Specific activity 

A Gas phase at start 
B "CO2 loop 
C Dissolved CO^, HCO3 
D NaH^COj injected 

C + D 












5.07 /iiClfimole 
4.95 fiClfimo\e 

' Effective volume. 

warmed at 60° for 10 min. After centrifugation and a further extraction with i ml of 
water, the combined clear extracts are concentrated at reduced pressure at below 
room temperature. The concentrated extract, or an aliquot portion thereof, is trans- 
ferred quantitatively to the paper chromatogram and analyzed in two dimensions 
(phenol-water, butanol-propionic acid-water) as in earlier work^. The location of 
the radioactive compounds on the chromatograph is found by radioautography 
with X-ray film. When necessary, overlapping phosphate esters are eluted, treated 
with phosphatase and rechromatographed. 

Determination of radioactivity in compounds 

The amounts of radiocarbon in each compound of interest on the chromatograms 
from each sample is measured with a Geiger-Mueller tube. The paper chromatogram 
is placed on top of the radioautograph, which rests on a horizontal light table, so 
that the darkened areas of the film may be seen through the paper. The Geiger- 
Mueller tube has a Mylar window, gold-sputtered for conductivity, but transparent 
and thin (less than i mg/cm^) to permit the passage of "C beta particles. This tube 
has an effective counting area of uniform sensitivity of about 17 cm^. The top of the 
tube is transparent plastic so that paper and radioautograph may be viewed through 
the top of the tube. Thus the counting area of the tube may readily be placed in posi- 
tion over the radioactive compound on the paper. If the radioactive area is more than 
4 cm across, or if it contains more than 20,000 counts/min (as counted by this tube 
on the paper), the radioactive area is divided into smaller areas which are counted 
one at a time (with the remainder of the spot covered by cards). The counting gas 
used is helium-isobutane (99: i). The counting voltage is about 1300 V. The sensitivity 
of the counter for '^C beta particles in an infinitely thin layer on an aluminum planchet 
is about one count/3.1 disintegrations. However, only about one-third of the beta 
particles escape from the paper (Whatman No. 4) and the actual sensitivity of this 
tube for ^*C in compounds on the paper is about 1/11.2. These sensitivities were 
determined by comparison of counts from three aliquot portions of a known ^■'C 
labeled solution : (a) chromatographed on paper, (b) dried on a planchet, and (c) placed 
in a scintillation counter with an internal standard. The radioactivity of each com- 
pound is counted on each side of the paper and an average is taken of the counts 
from the two sides. Comparison with determinations of radioactivity of compounds 
quantitatively eluted and placed on planchets indicates that this method of counting 
gives an accuracy of ± 5 %• 



Rate measurements 

Gas exchange: Measurements of the rates of COj uptake, ^*C uptake and O^ 
evolution by the photosynthesizing algae are made by taking the slopes of the three 
traces on the recorder. In order to obtain accurate readings in 10 min or less, the 
total effective gas volume of the closed circulating system is made small, about 
435 ml. With 80 ml of 0.5 % algal suspension in the system the resulting change in 
Oj or COg pressure is about 0.5 % in 10 min in a typical experiment. This corresponds 
to a rate of 22 /umoles of gas exchange/min/ml of wet packed algae. The response of 
the Beckman Infrared Analyzer, model 15 A, used in these experiments is not 
completely linear in the range used (0 to 2.0 °o CO,) so that a correction based on a 
previously obtained calibration curve is applied to the COg uptake curve plotted 
on the recorder. The response of the A. O. Beckman oxygen analyzer is essentially 
linear in the range used (19 to 21 °o). The level of '■*€ is plotted on the recorder as 
millivolts response of the Applied Physics Corpn.'s Vibrating Reed Electrometer to 
the ionization chamber (volume 118 ml, R = 10' ohms). From the known calibration 
of the ionization chamber this reading can be directly converted to /xC of "C. From 
the ^^COg reading and the ^*C reading the specific radioactivity of the CO2 may be 
calculated at all times during the experiment. This specific radioactivity is used to 
convert the rates of change of radioactivity in the system to rates of change of what 
we shall call ""C" throughout this paper. For convenience of expression and calcula- 
tion, this '^C will be expressed in ;umoles and represents the amount of ^^C and "C 
corresponding to a given measured amount of radioactivity in the CO 2 administered 
to the algae at any time during the experiment. 

Total fixation in algae: In some experiments, small aliquot portions of each 
sample of algal material, taken and killed in alcohol during the course of the experi- 
ment, are spread in a thin layer on planchets with acetic acid, dried, and counted. 
The amount of '^C found at each time of exposure of the algae to "CO 2 is plotted 
and the slope of the curve drawn through these points gives the rate of appearance 
of "C in stable compounds in the plant. 

Fixation of ^*C in compounds found on the paper chromatograph : After the '^C 
in individual compounds found on the paper chromatogram has been measured, the 
amounts are sometimes totaled for each sample up to one minute, and a rate of ap- 
pearance of "C in these compounds is calculated. 

Steady state Expt. 18 

The rates of exchange of gases before, during and at the end of the experiment 
are shown in Table IV. We shall take 15.5 ^moles/min as an average value for uptake 
of carbon during the experiment. 

Ahquot portions of the samples were dried on planchets and their radioactivity 
was counted. When results of these counts were plotted versus time of sampling, the 
rate of fixation of "C into compounds stable to drying on the planchets was found 
to be about 15 /xmoles/min (Fig. 3). 

After chromatographic separation of the compounds, radioautographs, of 
which Fig. 4 is typical, were obtained. The radioactivity of each compound in each 
sample was determined and the total radiocarbon found in the various compounds 





All rates are given in /tmoles/ml of wet packed algae. 

Carbon dtoxide 

Initial rate 17-9 

During experiment 16.6 15.1 

Final rate 141 i3-7 

* See section Methods of measurements of rate of gas exchange for explanation of expression of 
"C in //moles. In theory the value for "C and CO^ should be the same. The difference is a reflection 
of inaccuracy in measurement of the slopes, especially CO^. 

Fig. 3. Appearance of '^C in stable compounds 

(dried on planchets) in Chlorella pyrenoidosa vs. time 

of photosynthesis with "COj. 

JO 40 50 60 m 
Time in Seconds with"C02 

Fig. 4. Radioautograph of chromatogram of Chlorella 
pyrenoidosa after 2 min photosynthesis with "COj. 


\, titim tnjt, *»■ 

Fig- 4- 




was plotted against time (Fig. 5). The maximum slope of the curve in Fig. 5 is 13 
^moles. This is a lower limit for the rate of appearance of i*C in soluble compounds 
which are also stable to chromatography. It does not take into account other com- 
pounds, too weakly radioactive to be counted, or "lost" from the front of our chroma- 
tograms. (In order to obtain good separation of phosphate esters we customarily 
allow the phenol-water solvent to drip from the ends of the chromatograms. Small 
amounts of labeled fatty material are lost in this way.) 

After 30 sec, appreciable amounts of radioactivity are passing through the 
extractable precursor compounds seen on the chromatograms into non-extractable 
substances, which are not seen on the chromatograms. Consequently the rate of 
appearance of **C in compounds on the paper decrease. 

During the first ten seconds, the rate of appearance of i*C in stable compounds 
is less than the maximum rate during the subsequent time. This could be ascribed 
to mixing time of the added H^CO^with the H^^CO^ present initially, or alternately 
to the presence of an intermediate pool of either HCO3 or some other unstable or 



30 40 50 60 70 
Time in Seconds with ^^COg 

Fig. 5. Appearance of '•'C in compounds on chromatograms prepared from Chlorella pyrenoidosa vs. 

time of photosynthesis with ^^COj. 

volatile compound. Such a compound would precede the stable soluble compounds 
in the fixation pathway. The size of this "pool", if it exists, cannot be greater than 
the difference between the fixation curve after 10 sec and a line of the maximum slope 
drawn through the origin (see Figs. 3 and 5). This is no more than i.o to 1.5 /nmoles, 
which is equal to the carbon fixed in 4 to 6 sec in this experiment. A calculation of 
the amount of HCOJ which would be found inside algae cells in a volume of i ml with 
an internal pH of 7 in equihbrium with 1.7 % CO^ gives a value of about i to 1.5 
/j,moles, depending on the volume available inside the cells. It seems to us to be not 
unreasonable to suppose that this "pool" is merely intracellular COg and HCO3 
but it does not matter to the subsequent argument whether it is this or some other 
unstable or volatile substance. 

From the measured rates of uptake of COj and "C and from the rates of ap- 
pearance of i*C in stable compounds these experimental findings may be listed: 
(a) The appearance of ^*C in stable, nonvolatile compounds, after the first 10 sec 
of exposure of the plant to ^*C02, is equal to the rate of total uptake of "COj within 




experimental error, (b) During the period between lo and 30 sec exposure to "COj, 
the appearance of ^^C in individual compounds which can be isolated by our methods 
of paper chromatography, is equal to at least 85 % of the rate of total uptake, (c) If 
there is a pool of COj, HCOJ or other unstable or volatile compound lying between 
administered COj and stable compounds in the fixation pathway, its amount is not 
more than i.o to 1.5 /xmoles (4 to 6 sec fixation) and it is essentially saturated after 
10 sec. 

Let us next consider the question of how much of this fixed ^*C must pass through 
the PGA pool. 

In Fig. 6 are shown the labeling curves of some of the more rapidly labeled 
compounds and groups of compounds. By 3 min, compounds of the carbon reduction 
cycle are essentially saturated with radiocarbon. Secondary intermediates such as 
sucrose, malic acid, and several amino acids are not saturated until longer times 
(5 to 30 min). In order to evaluate the importance of the fixation pathway leading 
through PGA, we have tabulated the actual measurements of "C found in compounds 

Fig. 6. Appearance of '*C in PGA and sugar phosphates in Chlorella pyrenoidosa vs. time of photo- 
synthesis with '^COj. 

during the first minute (Table V). The "C found in all those compounds derived 
from PGA without further carboxylation (see Fig. i) is added together (Tj). Com- 
pounds labeled by C3-C1 carboxylation are totaled separately (Tj). Since three of 
the carbon atoms in these compounds are derived from PGA, their total radioactivity 
is multiplied by a factor which is 3/4 times the degree of saturation of the PGA, 
which is presumed to be the same as that of their immediate precursor, namely, PEPA. 
(The saturation curves for PGA and PEPA are in fact very similar in this and other 
experiments.) The sum of Tj and Tjf, representing measured '^C derived from the 
primary reaction which forms PGA, is plotted in Fig. 7. Again the "pool" of HCO3 
or other volatile or unstable compound is about i /^mole and in this case it must 
precede PGA in the chain of reactions. Where one draws the curve of maximum 
slope through these points is somewhat arbitrary, but the maximum rate of ap- 
pearance of '^C in these compounds falls somewhere between 11 and 13 /xmoles/min. 
Thus on the basis of the appearance of "C in these extractable, stable compounds 
alone, at least 70 to 85 % of all carbon fixed during photosynthesis (measured ex- 
ternally) is incorporated via the carbon reduction cycle. It must be emphasized 












21. M 




































(?» N o fo 

- O O 

6 lo d 6 

I I 

a ? A 



O (N moo Noo mrOy-jM ino 
po o "^ -H o o d o « ■-' fo w 

i i ? J AA 

O lOPOrOrO»00^0 

0\ t^ fn 10 fOGO O N O M 
N — r^O^'-'N N WW >« 

N d N 6 6 d d d d d d d 


H- d - d d d d d d d d d 

N 11 o -« r-.r^roioinroc^t o 
M d -< d d d d d d d d d 



n^t^r^t^iriTj-rrjr^iOrnoO 0> 

00 ^ i^ T^ moo to lO t--. >-i o >-" 

d d d d d d d d d d d d 

in^'N lO'N^O ■^■^TT — 



^ \0 lOO — OOvOOO ■^N 

d d d d d d d 





000 f*^r^ONpr>fO(N u-jiTi 
MOO(NO — '-■■-•Ni-i 


moo r^-i r-io ^T^ 
q q o q q o o o 
d d d d d d d d 

TO _C 

u: o 

S o 

2 E 

< O £ 

< CU X CI, 


1) j3 

(U c 
in X 

.. o ° 

<n x: 3 
.2 cxjo 

c . 






00 o> 



ro (N 


10 — 10 

d d d 

00 fO lO t-^ 
O «-< O W 






ro O O O 


00 vnoo n-i 

O^ M O O 

(<i q q o 

<N d d o 


C^ lO rrj o 

!S ►. O "1- 

vO O O O 


o o 

\0 r^ CI 00 

oi q q o 
d d d d 
















< fc s 

o u-j T^ o o o 

»^ O r^ O 00 ' 
r^ 10 u-)Oo 

o d d d 00 d 



"N u-i O O — O 

in m r^ -. o m 

■^o -^ (N m M 

6 6 6 6 ^ d 

O u") FN 10 r^ 10 
— r--- m m TT r^ 
m 10 -^ ■ 


d d 





- r^oo 






U^ IT) 
M — 


q vD 















« N X 1 I 
H X ~ H O H 




that this percentage is a lower limit based only on absolute measurements of identified 

A lower limit for the amount of carbon incorporated via C^ plus Cj carboxylation 
is obtained by plotting Tg — ^Tjf (Fig. 7). The minimum rate of this incorporation is 
about 0.4 /imoles/min/ml algae, or about 3 % of the total. Note that this value is 


to 20 30 ■W 50 60 

Time in Seconds with ^^C02 

Fig. 7. Appearance of "C in compounds derived from PGA and in compounds derived from C, + Cj 
carboxylation in Chlorella pyrenoidosa vs. time of photosynthesis with "COj. 

for the actual introduction of COj and does not include the carbon derived from PGA 
(Tgf). The rate of incorporation of '*C into these three compounds thus accounts for 
about 4 times 3, or 12 % of the total in this experiment. Other experiments indicate 
that the relative contribution of C3-C1 carboxylation varies considerably and tends 
to be higher (up to 3 times that reported in this case) when the rate of COg fixation is 
greater and when amino acid synthesis is more rapid. In addition to the three com- 
pounds listed here, other substances may be derived in part from C^-Cg carboxylation, 
such as glutamic acid and citric acid, discussed below. 

While at least 73 % of the total rate of fixation of carbon has thus been shown 
to be due to the carbon reduction cycle and C1-C3 addition, there is no indication 
of any other significant fixation pathway. In Fig. 8 the ^*C found in glutamic acid 
and in citric acid is shown. Could this labeling of glutamic acid be the result of a 
carboxylation of y-aminobutyric acid? The maximum rate of labeling of glutamic 
acid and in citric acid is shown. The maximum rate of labeling of glutamic acid is 
about 0.7 /^moles/min or 4.5 % of all ^*C fixed. Since this rate is found between 5 and 
20 min, it probably represents labeling of all five carbon atoms of glutamic acid, 
because the precursors are surely at least partially labeled after 5 min. The labeling 
due to carboxylation reaction would be expected to begin during the first 30 sec, if 
one is to judge by the other known carboxylation reactions which were discussed 
earlier. Yet, after the first 31.5 sec, the glutamic acid contains only 0.02 fimoles of 
i*C. Between 40 and 60 sec, its labeling rate is only 0.2 /xmoles/min. Moreover, 
y-aminobutyric acid itself would have to be synthesized from CO2 (by some as yet 




unknown route), if it were a precursor to glutamic acid, and would have to be ap- 
preciably labeled by the time glutamic acid reaches its maximum labeling rate. Yet 
ije can detect no radiocarbon in y-aminobutyric acid in this experiment or in others 
fy this series, even after the algae have been exposed to ^^COj for lo min. Clearly, 
wottle if any of the labeled glutamic acid formed in our experiments is made b 

^ — i—i — i — * — > — * — ir-io 

Tims in Minuttt 

Fig. 8. Appearance of ^*C in PGA, glutamic acid and citric acid in Chlorella pyrenoidosa vs. time of 

photosynthesis with "COj. 

carboxylation of y-aminobutyric acid. Rather, it must arise from other intermediate 
substances such as those formed by the two carboxylation mechanisms already 

Note, however, that the rate of labeling of citric acid is by far too small to permit 
it to be the precursor of the labeled glutamic acid in any sequence such as : 
oxaloacetic acid + acetyl coenzyme A — >■ citric acid ->■->■->■ 
a-oxoglutaric acid -|- COj — > glutamic acid 

Steady state Expt. 28 

All the results described thus far were obtained in an experiment (steady state 18) 
in which the nutrient solution, though not automatically replenished, was sufficient 
to maintain the rate of photosynthesis at a nearly constant level during the course 
of the experiment. The results of steady state Expt. 28, in which the nutrient solution 
was replenished during the course of the experiment led to the same conclusions. 




COi uptake 



Rale of appearance 

cf "C in compounds on 


(20-40 sec) 


at 40 sec 

PGA residual' carbon 
saturalion according to 

Reaction D Reaction L 








* See subsequent discussion for explanation of the term "residual". The degree of saturation 
at 40 sec is obtained by dividing the measured value of '*C in the compound at 40 sec by the 
saturation level of **C in the compound (or residual atoms) after 10 min exposure of the algae 
to "CO,. 



These results are summarized and compared with steady state Expt. 18 in Table VI. 
Though not shown in the table, the maximum rate of appearance of ^*C in observable 
compounds derived from the carboxylation reaction leading to PGA (the carbon 
reduction cycle) was 70 to 90 % of the externally measured rate of i*C uptake. 


When Calvin and Massini" reported the formation of PGA in an overall reaction 
requiring ribulose diphosphate and CO 2 they proposed that the reaction in the light 
gave one molecule of PGA and one of triose phosphate but in the dark gave two 
molecules of PGA. Wilson^* discussed this possibility further after it was realized 
that the carboxylation did not involve an intermediate splitting of the ribulose to 
triose and diose. The dark reaction in whole plants'^ and the reaction in isolated 
enzyme systems^*-!' was found to give rise to two PGA molecules. Also, it is clear 
from previous kinetic studies^- ^^ of carbon fixation during photosynthesis that the ^^C 
entering the carbon reduction cycle via the ribulose carboxylation passes through 
the carboxyl group of PGA initially. This is consistent with the fact, established for 
the isolated enzyme system by Horecker'*, that the CO 2 is bonded to the number 
two carbon atom of ribulose diphosphate. More recently Park^' has shown by means 
of inhibition studies in broken spinach chloroplasts that "C entering that system 
must pass through PGA. That is, PGA is a biochemical intermediate compound — not 
merely a compound formed by thermal breakdown after the plant is killed. 

We shall present here an argument, based on kinetic data, which indicates that 
the carboxylation of RuDP in vivo during photosynthesis gives rise to only one 
molecule of 3-PGA. 

If the i*C which has just entered PGA from "CO2 is subtracted from the total "C 
in PGA, the i*C in the remaining carbon atoms of the PGA must all be derived 
from ribulose diphosphate. 

Let us consider the two reactions : 




2 1 

c = 








HCOH + ♦COj -^ 






bI or L) 







— c— 





— c— 


The position of the '*C which has just entered the cycle as "CO^ is indicated by the 
asterisks. In reaction D, there are five remaining carbon atoms of PGA (numbers i 
to 5) which must be derived from RuDP, while in reaction L there are two such 
"residual" carbon atoms (numbers i and 2). The steady state concentration of PGA 
in steady state Expt. 18 is 3.0 /xmoles of carbon/ml algae, hence the carboxyl carbon 
concentration is i!o jumole of carbon. However, if reaction D is correct, only one-half 



of this carboxyl carbon, or 0.5 /xmole, is derived immediately from CO^; the other 
half (carbon atom 3) comes from RuDP. We shall subtract the "C due to newly 
incorporated "COg from the total i^C found in PGA at each time and for each of these 
two cases. The specific radioactivity of the remainder may then be compared with 
the specific radioactivity of the RuDP from which it must be derived. 

In order to make this subtraction it is necessary first to calculate the radiocarbon 
in the carboxyl group of PGA as a function of the time of exposure of the algae to 
"COg. This calculation requires in turn a calculation of the saturation curve of the 
"CO 2 pool", although this could be assumed to be saturated from the beginning 
without seriously affecting the results. 

Consider the steady state system : 

R R 

CO2 — > Pool 1 — > Pool 2 — >■ etc. 

Let Ci and C^ be the steady state concentrations of Pools i and 2 and let x and v 
be the degrees of saturation with "C of these pools (respectively) as a function of 
time of exposure of the algae to "COj. R is the rate of flow of carbon into the system 
and through the two pools. It is also assumed in this case that the rates of the back 
reactions are negligible compared to the rates of the forward reactions. 

For a small increment of time, the change in degree of saturation is the difference 
between the rate of flow of "C into the pool (R) and the rate of flow of carbon out of 
the pool (Rx), divided by the size of the pool C^; dxidt = (R—Rx)IC^. Integration 
and determination of the integration constant at / = gives x = i — expt ( — R/Ci)0 . 

During a small increment of time, the change in degree of saturation of the second 
pool is the difference between the rate of flow of "C into the second pool {Rx) and the 
rate of flow out [Ry) divided by the pool size C^; 

Integration and determination of constants at i = o leads to two solutions, one for 
the case Cj + C^- 

and another for the case C^ = C^'. 

y = I — (I — RtjC) exp (— RtjC) 

In applying these equations to the data from steady state Expt. 18 we have assumed 
a value of Cj = 1.2 /xmoles for the "COg pool" (Fig. i) and a value of 0.2 /^moles/sec 
(= 12 /ixmoles/min) for R. The resulting values for x are shown by curve A, Fig. 9. 

If reaction D is correct, the PGA carboxyl pool arising from newly incorporated 
CO2 is 0.5 /Limoles and its degree of saturation jy is given by curve B, Fig. 9. If reaction 
L is correct, this pool is i.o /xmole and the saturation curve y is that shown as curve C. 
Curve B times 0.5 and curve C times i.o give, as a function of time, the respective 
/xmoles of "C in the PGA carboxyl pool derived directly from COj. 

The degree of saturation of the residual carbon atoms of PGA (those which are 
derived from RuDP) may now be calculated by subtracting from the experimentally 
determined ["C]PGA these values of the COa-derived carboxyl (0.5 S for reaction D, 
1.0 C for reaction L) and dividing by the pool sizes of the residual carbons (2.5 and 




2.0 respectively). The resulting saturation curves are shown in Fig. 10. In the same 
figure, Curve R is the saturation curve for ribulose diphosphate, obtained by dividing 
the experimentally determined i*C labeling of RuDP by its steady state concentration, 
which was 0.36 ^moles/ml algae. 

If the carboxylation of RuDP were to lead to the formation of two molecules 
of PGA (reaction D), then all of the carbon atoms of RuDP must give rise to the 
"residual" carbon atoms of PGA. The degree of saturation of these residual carbon 
atoms at no time could exceed the degree of saturation of the carbon atoms of RuDP. 
Since the calculated values for these residual atoms, (PGA-0.5 B)/2.5, do exceed 
those of RuDP at all times after 12 sec, reaction D does not appear to be correct. 
The curve for reaction L does not exceed the saturation of RuDP until about i min. 
In this case, the residual carbon atoms of PGA are derived only from carbon atoms 2 
and 3 of RuDP, and thus may exceed the saturation of the average of carbon atoms 

Time in Seconds 

"T5 'X 30 40 50 eo' 
Time in Seconds 

ao 90 100 

Fig. 9. Degree of saturation (vs. time of photo- 
synthesis with "CO2) of "CO2 pool" and of 
PGA carboxyl derived immediately from "COj 
according to two proposed carboxylation re- 
actions. Curve A is for "COj pool", curve B is 
for PGA carboxyl derived immediately from 
^^COj according to reaction D, curve C is for 
PGA carboxvl according to reaction L. 

Fig. 10. Degree of saturation of ribulose di- 
phosphate (R) vs. time of photosynthesis with 
"CO2 compared with degrees of saturation of 
residual carbon atoms of PGA according to two 
proposed carboxylation reactions. 

I, 2, 3, 4, and 5 of RuDP. In fact, this is not surprising, since earher degradation 
studies on RuDP' showed that, during i*C incorporation in photosynthesis, carbon 
atom 3 is first labeled, followed by carbon atoms i and 2, followed finally by carbon 
atoms 4 and 5. The saturation curve for the residual PGA carbon atoms according 
to reaction L is thus about as would be expected. 

Note that after 30 sec the carboxyl carbon of PGA would be saturated and the 
same conclusion could be reached by looking only at the curves from 30 to 90 sec, 
which are not dependent on the foregoing calculations of CO^ pool and PGA carboxyl 
saturation. At these longer times it is sufficient to plot simply the curves for (PGA-0.5)/ 
2.5, (PGA-i.o)/2.o, and RuDP/0.32 all as a function of time. 

We conclude, therefore, that the labeling curves for PGA and RuDP in this 
experiment can best be interpreted as resulting from the occurrence of reaction L. 
That is, the in vivo carboxylation reaction of the carbon reduction cycle during 



photosynthesis appears to produce one molecule of PGA and one molecule of some 
other three carbon compound. 

Steady state Expt. 28 gav^e very similar results, from 10 sec to saturation (see 
Table VI for comparison at 40 sec). 

From these experiments alone we cannot identify this three carbon compound. 
It could be merely a small pool of PGA itself, tightly bound to an enzyme, or in some 
other way kept apart from the principal PGA pool. Such a pool of PGA molecules, 
if sufficiently small (> o.i /xmole), would not be distinguishable from the other 
PGA pool by our methods. 

Alternatively, the six carbon product of the carboxylation reaction may be 
reductively split to one molecule of 3-PGA and one molecule of triose phosphate. 
In either case, the requirement for the reaction leading to PGA and triose phosphate 
must be light (or cofactors derived from the light reaction), and the intact chloroplast, 
or some intact sub-unit of the chloroplast, as it occurs naturally in the living cell. 

One cannot say at the present time whether or not any of the chloroplasts or 
chloroplast fragments isolated from broken cells retain the capacity to carry out such 
a reductive splitting of the six carbon intermediate of the carbon reduction cycle. In 
such cell-free systems, the carbon reduction cycle may well operate only via the 
carboxylation reaction leading to two molecules of free 3-PGA. Recently Park^" 
has prepared electron micrographs of chloroplast and chloroplast fragments which 
had been found by him to have about as high a rate of photosynthetic COg reduction 
as any such rates reported for cell-free systems. When compared with electron micro- 
graphs of chloroplasts in intact cells, these isolated fragments appear to have under- 
gone considerable physical change, particularly in regard to the apparent density 
of the stroma and spacing between lamellae. It is possible that the reductive carboxy- 
lation pathway, if correct, operates only in the unaltered lamellar system by means of 
some rather direct transfer of photochemically-produced reducing power from the 
pigmented layer to the carbon reduction cycle. 

If two different three carbon compounds are formed in vivo in the light by the 
carboxylation of RuDP, and if these two products are kept separate until they have 
been converted to triose phosphate, and react with each other to give hexose, then 
the resulting hexose molecule might be dissimilarly labeled in its two halves, nameyl 
carbon atoms i, 2, and 3, and carbon atoms 4, 5, and 6. Such asymmetry has been 
reported by Gibbs and Kandler^^-^^. However, other explanations of the phenom- 
enon are also consistent with the carbon reduction cycle^. 


The work described in this paper was sponsored by the United States Atomic Energy 
Commission, University of California, Berkeley, Calif. (U.S.A.). 


1 J. A. Bassham, a. a. Benson, L. D. Kay, A. Z. Harris, A. T. Wilson and M. Calvin, /. Am. 

Chetn. Soc, 76 (1954) 1760. 
* M. Calvin, /. Chem. Soc, {1956) 1895. 
' J. A. Bassham and M. Calvin, The Path of Carbon in Photosynthesis, Prentice-Hall, Englewood 

Cliffs, New Jersey, 1957. 



* M. Calvin and A. A. Benson, Science, 109 (1949) 140- 

' A. A. Benson, J. A. Bassham, M. Calvin, T. C. Goodale, V. A. Haas and W. Stepka, /. Am. 
Chem. Soc, 72 (1950) 1710. 

• A. A. Benson, Arch. Biochem. Biophys.. 32 (1951) 223- 
' A. A. Benson, /. Am. Chem. Soc, 73 {1951) 2971- 

» A. A. Benson, J. A. Bassham, M. Calvin. A. G. Hall, H. E. Hirsh, S. Kawaguchi, V. Lynch 
AND N. E. ToLBERT, /. Biol. Chem., 196 (1952) 7°i- 

» M. Goodman, D. F. Bradley and M. Calvin, /. Am. Chem. Soc, 75 (i953) 1962. 

"" M. Goodman, A. A. Benson and M. Calvin, /. Am. Chem. Soc, 77 (i955) 4257- 

" O. Warburg, Science, 128 (1958) 68. 

" R. W. Krauss, in J. S. Burlew, Algal Culture from Laboratory to Pilot Plant, Carnegie Institu- 
tion of Washington Publication 600, Washington D.C., 1953, p. 94- 

" M. Calvin and P. Massini, Experientia, 8 (1952) 445- 

" A. T. Wilson and M. Calvin, /. Am. Chem. Soc, 77 (i955) 5948- 

" J. A. Bassham, K. Shibata, K. Steenberg, J. Bourdon and M. Calvin, /. Am. Chem. Soc, 
78 (1956) 4120. 

*• A. Weissbach, B. L. Horecker and J. Hurwitz, /. Biol. Chem., 218 (1956) 795- 

I' W. B. Jakoby, D. O. Brummond and S. Ochoa, /. Biol. Chem., 218 (1956) 811. 

w A. A. Benson. S. Kawaguchi, P. M. Hayes and M. Calvin, J.Am. Chem. Soc, 74 (1952) 4477- 

" R. B. Park. N. G. Pon. K. P. Louwrier and M. Calvin, Biochim. Biophys. Acta, 42 (i960) 27. 

»" R. B. Park, Annual Winter Meeting, The Western Society of Naturalists, Los Angeles, University 
of Southern California, December 1959. 

" M. GiBBS and O. Kandler, Plant Physiol., 31 (1956) 4ii- 

»« M. Gibes and O. Handler, Proc Natl. Acad. Sci. U.S., 43 (i957) 446. 

Biochim. Biophys. Acta, 43 (i960) 447-464 





Acetic acid 

as precursor to glutamic 

acid, 35 
fat formation from, 57 
formation of, 39, 47 
from malate?, 38 
Acetoacetyl CoA, 57, 60 
Acetyl CoA, 38, 57 

formation from pyruvate, 47 
Acetyl phosphate, 48 

formation from sugar phos- 
phates, 41 
Adenosine triphosphate (ATP) 
free energy of formation, 4 
function, 5, 6, 17 

an early product of photo- 
synthesis, 29 
mechanism of formation, 33, 

in preillumination experi- 
ments, 17 
pyruvate formation from, 47 

rate of formation in Chlo- 
rella pyrenoidosa, 31 

two pools of, 30 
Aldolase, 9, 16,51 
Algae, 12 
Amino acids, 25 

synthesis of, 29 
y-Aminobutyric acid, 34 
8-Amino levulinic acid, 64 
Aromatic compounds, 65 
Aspartic acid, 17,29,30,31 
ATP {see Adenosine triphos- 

Carbohydrates, 25 

Carbon reduction cycle of 
cofactor requirements, 9 
description of, 8-12 
evidence for, 12-16 
figure, 11 


2-Carboxy-3-keto- 1 ,5-diphos- 

phoribitol, 1 1 
Carboxydismutase, 23, 51 
Carboxyl carbon of PGA, 39 
Carboxylating enzyme, 51 
Carboxylation reactions 
of 8-aminobutyric acid, 34 
balance among, 26 
evidence for RuDP as react- 
ant and PGA as product, 
keto acid, product of, 8 
leading to C4 compounds, 22 
number of times per cycle, 9 
of phosphoenolpyruvic acid, 

other carboxylations than 

cyclic, 21 
PGA, first stable product of, 

quantitative importance of, 

reductive, 1 1, 23 
Carboxylic acids 

formation of, 37, 38, 39 
Carotenoids, 60 
Chlorella pyrenoidosa, 12, 25, 

Chlorophylls, 60, 64 

amino acid pools in, 30 
as biosynthetic factory, 68 
compounds excreted by, 28 
site of synthetic reactions, 27 
transport of reducing power 
from, 40 
Citramalate, 35 
Citric acid, 48 
Citrulline, 31 


of glycolic acid, 39 

of hexoses, 16 

of PGA, 15,39 
Dihydroxyacetone phosphate 

early product of CO2 reduc- 
tion, 49 

glycerol phosphate forma- 
tion from, 59 

intermediate in carbon re- 
duction cycle, 1 1 

reactions of, 51 

Epimerase, 9 

Erythrose-4-phosphate, 11,49, 

Farnesyl pyrophosphate, 61 

Fatty acids, 25, 56 

Fatty acid synthesis, 38 

Fats, 56, 67 

Free energy change 

carboxylation reaction, 38 

formation of ATP, 4 

formation of TPNH, 4 

relation to reversibility, 38 
Fructose- 1 ,6-diphosphate 

in carbon reduction cycle, 11, 

reactions of, 51 
Fructose-1-phosphate, 55 

in carbon reduction cycle, 1 1, 

reactions of, 51, 53, 55 


Fructose-6-phosphate ketolase, 

Fumaric acid, 37 

Galactose, 49 

Geranylgeranyl pyrophosphate, 

Glucose- 1,6-diphosphate, 52 
Glucose- 1 -phosphate, 49, 54 
Glucose-6-phosphate, 49, 52 
Glutamic acid 

formation of, 34, 35 
light-dark labeling of, 48 
rate of formation, 31 
two pools of, 30 
Glutamine, 31 

a-D-Glyceryl-1-phosphate, 59 

as porphyrin precursor, 62 
origin of, 36, 38 
rate of formation, 31 
slow labeling of, 32 
Glycolaldehyde-thiamine pyro- 
phosphate, 41 
Glycolic acid 

direct formation from CO2, 

effect of CO2 level on forma- 
tion, 40 
effect of Mn+ + deficiency 

on, 46 
formation in barley seed- 
lings, 39 
formation from sugar phos- 
phates, 43 

labeling by T and C^*, 45 
labeling of carbon atoms, 39 
role in hydrogen transport, 46 

Glycolyl CoA, 43, 44 

Glyoxylate, 44, 62 

Glyoxylate cycle, 48 

Glyoxylic acid, 38, 39 

Hematin pigments, 60, 64 
Hexoses, degradation of, 16 
/3-Hydroxy-/3-methyl glutaryl, 

•y-Hydroxyglutamic acid, 35 

Inorganic phosphate, 51 
Isomerase, 9 

Isopentenyl pyrophosphate, 61 
Isoprene unit, 61 


acid-7-phosphate, 65 
Ketoglutaric acid, 33 

Malic acid 

early fixation product, 17, 29 

formation of, 37 

reactions of, 38, 63 
Malonyl CoA, 57 
Mannose, 49 

Mannose-6-phosphate, 53 
Mesaconic acid, 35 


^-Methylaspartate, 35 
Mevalonic acid, 61 
Monosaccharides, 49 

Oxalacetic acid, 1 1 

Paper chromatography, 12 
PGA, 19, 20, 30 

degradation of, 15, 39 
Phenylalanine, 65 
Phosphatase, 51 
Phosphoenolpyruvic acid 
in amino acid synthesis, 30, 

in aromatic synthesis, 65 
carboxylation of, 26 
formation from PGA, 17 
formed from carbon reduc- 
tion cycle, 1 1 
Phosphoglucomutase, 52 
3-Phosphoglyceraldehyde, 9, 49 
2-Phosphoglyceric acid, 1 1 
3-Phosphoglyceric acid (3- 
PGA), 11, 15 
product of carboxylation re- 
action, 9, 11 

Phosphoglycerylkinase, 9, 51 
Phosphohexose isomerase, 52 
Phosphoketolase, 41, 48 
Phosphoribulokinase, 51 
Phosphoroclastic cleavage, 41 
Phosphorylated sugars {see 
Sugar phosphates) 

Phosphoshikimic, 35 

Phytoene, 62 

Phytol, 60, 61 

Pigments, 60, 67 

Polyglycerol phosphates, 59 

Polysaccharides, 67 

Porphobilinogen, 64 

Porphyrin compounds, 62 

Protein synthesis, 29 

Proteins, 29, 67 

Protoporphyrin 9, 64 

Pyrophosphate, 52 

Pyruvic acid 

from glutamic acid, 35 
oxidative decarboxylation 

of, 47 
transamination of, 34 


of photosynthesis experi- 
ments, 15, 18 
preparation of, 14 
use of, 12 
Radiocarbon (C^^), 12, 13 

in carbon reduction cycle, 11, 

reactions of, 51 
Ribulose-l,5-diphosphate (Ri- 
bulose diphosphate) 
in carbon reduction cycle, 11, 

carboxylation of, 20, 51 
decrease in light-dark experi- 
ment, 19 


Ribulose-5-phosphate, 9, 11, 49 
RNA, 67 

Sedoheptulose- 1 ,7-diphosphate, 

Sedoheptulose-7-phosphate, 49, 

Serine, 30, 31 

Shikimic acid, 65 

Squalene, 61 

Steady-state growth, 21 

Steroids, 61 

Succinate, 62 

Sucrose, 53 

Sucrose phosphate, 52, 54 

Sucrose phosphatase, 52 

Sugar phosphates 

conversion to free sugars, 50 

occurrence as photosynthetic 

intermediates, 15, 16, 49 

Terpene compounds, 60 
Tetrahydrophytoene, 62 
Thiamine pyrophosphate, 41 
Threonine, 31, 36 
TPN, TPNH {see Triphospho- 

pyridine nucleotide) 
Transaldolase, 16 
Transaminase, 34 
Transketolase, 9, 16, 51 
Triosephosphate dehydrogen- 
ase, 9, 51 

Triosephosphate isomerase, 51 
Triphosphopyridine nucleotide 
free energy of formation, 4 
function, 4-6, 17 
requirement in cycle, 9, 17 
requirement for RuDP for- 
mation, 19 
Two-dimensional paper chro- 
matography, 13 
Tyrosine, 65 

UDPG-4-Epimerase (Galacto- 

waldenase), 52 

transglycosylase, 52 
UDPG-pyrophosphorylase, 52, 

Uridine diphosphogalactose, 52 
Uridine diphosphoglucose, 52, 

Uridine triphosphate, 52 
Uroporphyrinogen, 64 

X-ray film, 13 

in carbon reduction cycle, 11, 

phosphoroclastic split of, 41 

reactions of, 51 
Xylulose isomerase, 51